Topic:"Forecasting Methodology" in M11108

REDACTED (CONFIDENTIAL INFORMATION REMOVED) Nova Scotia Utility and Review Board IN THE MATTER OF The Public Utilities Act, R.S.N.S. 1989, c.380, as amended 2023 Load Forecast Report April 28, 2023 REDACTED REDACTED (CONFIDENTIAL INFORMATI...

AI summary The document outlines the 2023 Load Forecast Report prepared under the Public Utilities Act, R.S.N.S. 1989, c.380, as amended. It includes sections on methodology, major inputs, and forecasting approaches for energy demand in Nova Scotia.

1 List of Figures 2 3 Figure 1: Historical and Predicted Annual Net System Requirement ............................................ 7 4 Figure 2: Historical and Predicted Annual System Peak ....................................................

AI summary The document lists figures related to historical and predicted energy system requirements, peak demand, heating/cooling degree day trends, temperature regression models, and geographic weather station data. These visualizations support forecasting methodologies and energy usage pattern analysis for system reliability planning.

DATE: April 28, 2023 Page 3 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary A redacted section from the 2023 Load Forecast Report, dated April 28, 2023, page 3 of 98. The document contains confidential information removed, with no visible content beyond the title and page reference.

1 Figure 43: Illustrative Contribution of Specific End Uses............................................................ 61 2 Figure 44: Commercial Class Sales ...................................................................................

AI summary The text lists figures illustrating energy sales, demand forecasts, temperature regression models, and economic indicators. It includes historical and projected data for residential, commercial, and industrial sectors, along with demand response and peak temperature analysis.

57: Hourly Peak Temperature Regression Model Results ................................................ 80 16 Figure 58: 12hr Avg Lag Peak Temperature Regression Model Results ..................................... 80 17 Figure 59: 24hr Avg L...

AI summary The text lists figures analyzing peak temperature regression models, load research data, and forecast accuracy for energy systems. It includes historical and projected system peaks, demand response impacts, and residential/commercial end-use contributions, relevant to energy load forecasting and system reliability analysis.

Figure 72: Monthly historical Residential LRS load at peak and forecasts .................................. 93 31 Figure 73: AMI Peak Estimates ..................................................................................................

AI summary The document outlines the 2023 Load Forecast Report, including attachments with residential and commercial demand models, forecast classes, and appendices covering NS Power forecasts, model details, comparisons, and stakeholder presentations. Most content is redacted, with figures and appendices listed but not detailed.

Confidential) 20 Appendix E: Stakeholder Presentation DATE: April 28, 2023 Page 5 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary Confidential stakeholder presentation titled '2023 Load Forecast Report' from April 28, 2023, with content redacted. The document is part of a regulatory proceeding in Nova Scotia, focusing on energy load forecasting.

Page 5 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary The 2023 Load Forecast Report addresses electricity demand projections for Nova Scotia, likely involving analysis of heating/cooling degree days (HDD/CDD) and regulatory considerations under the Regulations of Nova Scotia (R.S.N.S.). The redacted content suggests technical energy planning discussions.

1 1.0 EXECUTIVE SUMMARY 2 3 In accordance with the Nova Scotia Wholesale Electricity and Renewable to Retail Market 4 Rules, Nova Scotia Power Incorporated (NS Power, the Company) is required to provide 5 the Nova Scotia Utility and Review...

AI summary NS Power must submit a 10-year load forecast to NSUARB, covering 2023-2033. The forecast considers sales history, weather, economic factors, and uses SAE models. Uncertainties from variables like weather and policy changes are acknowledged.

residential and commercial rate classes. The SAE models explicitly 27 incorporate end-use energy intensity projections into the Load Forecast. End-use energy 28 forecasts derived from the residential and commercial SAE models are then comb...

AI summary The 2023 Load Forecast Report projects a 0.7% annual increase in Net System Requirement (NSR), driven by new customer additions, EV adoption, and RTR market reductions. Long-term growth is tempered by DSM initiatives, efficiency gains, and solar installations.

erage annual increase of 0.7 14 percent. Annual historic and forecast NSR are shown below in Figure 1. 15 16 Figure 1: Historical and Predicted Annual Net System Requirement 17 18 DATE: April 28, 2023 Page 7 of 98 REDACTED (CONFIDENTIAL IN...

AI summary NS Power's 2023 Load Forecast Report projects a 0.7% annual increase in Net System Requirement (NSR) and a 2.3% annual rise in system peak demand, driven by customer growth, electrification, and EV adoption. Demand Side Management (DSM) and Demand Response (DR) programs are expected to mitigate some of this growth.

1 2.0 INTRODUCTION 2 3 NS Power annually develops a forecast of energy sales and peak demand requirements to 4 assess the effects of end-use and economic factors on the future power system load and 5 load shape. The forecast is a foundatio...

AI summary NS Power's 2022 Load Forecast Report, covering 2023-2033, was reviewed by the NSUARB through a paper hearing. Intervenors included the Consumer Advocate, Small Business Advocate, Industrial Group, EfficiencyOne, and Heritage Gas. The Board directed NS Power to implement agreed-upon recommendations, emphasizing improved detail on electrification impacts.

1 The Board directs NS Power to evaluate the model’s economic inputs, 2 including the COVID-19 variable and the assumptions and calculations 3 used to assess the impact of DSM, Solar PV, EVs, battery storage, and 4 weather. 5 6 The Board a...

AI summary The NSUARB directs NS Power to evaluate economic inputs in its load forecast model, including COVID-19 impacts, DSM, Solar PV, EVs, battery storage, and weather. It emphasizes stakeholder engagement and recommends refining elasticity assumptions, residential model variables, and housing completion data to improve forecast accuracy.

e in the province, re-evaluate the 29 use of housing completions for the near-term; 30 31 • Given the current inflationary environment, evaluate the use 32 of Median Household Income in place of Total Household 33 Income, as it is biased b...

AI summary The document outlines three actions for improving load forecasting: re-evaluating housing completions, using median household income over total income to address inflationary biases, and incorporating household demographics (size, age) to refine demand patterns by time of day. Data sources include the Conference Board of Canada and Statistics Canada.

1 - Revisit the short-term economic inputs provided by the Conference 2 Board of Canada to ensure data are close to those used in the 3 forecasts of Canada’s major Banks. 4 5 - Revisit the model’s EV adoption rates and examine EV rebates t...

AI summary NS Power revised the 2023 Load Forecast by updating peak temperature models, incorporating EV adoption data aligned with federal ZEV mandates, and including hybrid electrification scenarios. The Board directed revisiting economic inputs from the Conference Board of Canada and EV rebate data from Statistics Canada and Nova Scotia Open Data.

ng forecasting work. 19 20 In addition to the foregoing, NS Power also responded to questions from stakeholders 21 regarding the new all-time system peak set in February of 2023. 22 DATE: April 28, 2023 Page 13 of 98 REDACTED (CONFIDENTIAL...

AI summary NS Power addressed stakeholder inquiries about the new all-time system peak recorded in February 2023 and provided updates on load forecasting efforts. The 2023 Load Forecast Report, though redacted, is referenced as part of the proceedings.

1 3.0 FORECASTING APPROACH 2 3 NS Power continues to use a set of SAE models for the Residential 5 and Commercial 6 rate 4 classes, an econometric model for the Small and Medium Industrial classes, and customer 5 surveys and historical dat...

AI summary NS Power employs SAE models for residential and commercial classes, econometric models for small/medium industrial classes, and customer surveys/historical data for large industrial classes. The SAE model combines econometric and end-use methods, incorporating factors like efficiency trends, population changes, and weather. Structural changes influence long-term energy growth, as depicted in Figure 4.

ural changes are captured in 22 the residential forecast model through the SAE model specifications. Figure 4 shows the 23 general forecast approach used in the SAE models. 24 5 References to the Residential class include Domestic Service...

AI summary The document references the use of the SAE model specifications in the residential forecast model and mentions Figure 4 illustrating the general forecast approach. Much of the content is redacted, limiting detailed analysis.

1 4.0 DISCUSSION OF MAJOR INPUTS 2 3 4.1 Historical Class Sales and Energy Data 4 5 The Load Forecast is developed using NS Power’s “billed” sales rather than “accrued” 6 sales. Billed sales refer to the amount of energy billed to customer...

AI summary The Load Forecast uses 'billed' sales data for residential, commercial, and industrial sectors, with varying historical periods. The Renewable to Retail (RtR) market, established in 2016, allows licensed retailers to sell renewable energy directly to customers, with a third-party application pending approval by NSUARB.

e 16 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED 1 of energy through wind production, with 10 percent serving residential customers, 50 2 percent serving commercial customers, 30 percent serving ind...

AI summary The 2023 Load Forecast Report details energy distribution by customer type (10% residential, 50% commercial, 30% industrial, 10% losses) and explains weather data's impact on electricity sales via HDD/CDD metrics. Peak forecasts remain unchanged as NS Power must serve full peak demand regardless of energy source allocation.

od 19 January 2013 to December 2022. The average temperature continues to show a warming 20 trend: the 30-year average annual HDD is 3,892 while the 10-year average is 3,782. 21 NSUARB that if no electricity is sold to a customer under the...

AI summary The 2023 Load Forecast Report analyzes climate trends impacting Heating Degree Days (HDD) and Cooling Degree Days (CDD), showing a decline in HDD (-17/year) and increase in CDD (+1.4/year) due to warming. The NSUARB requires licensed retail suppliers to demonstrate justification for not selling electricity by 2024 (M10293).

Figure 7: HDD Trend 8 9 10 DATE: April 28, 2023 Page 19 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED 1 Figure 8: CDD Trend 2 3 4 5 These trends are reduced over time (approximately 40 years) such tha...

AI summary The 2023 Load Forecast Report analyzes HDD and CDD trends, projecting a 40-year reduction in annual HDD to 3,638 by 2033 and a rise in CDD to 124. This shift implies reduced winter heating demand and increased summer cooling demand for residential and commercial sectors, with leap year anomalies noted in 2024, 2028, and 2032.

𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 = �� � 𝑛𝑛 𝐴𝐴𝑡𝑡 𝑡𝑡=1 8 Where the 2022 monthly series (n=12) will be compared versus its respective forecast, At 9 is the actual value, and Ft is the forecast value. In both cases, forecast models have been 10 informed by 10 years o...

AI summary The document discusses the 2023 Load Forecast Report, comparing the accuracy of forecasts using one station versus multiple stations. It references the use of MAPE (Mean Absolute Percentage Error) to evaluate forecast performance, with a small difference noted between the two methods.

of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED 1 As the results above show, the differences are minimal. As no significant impact was noted, 2 this load-weighted approach was not incorporated into the...

AI summary The 2023 Load Forecast Report discusses the use of economic data, including median income and employment income, in forecasting load demand. The report notes that median income was evaluated as an alternative to household income, though data is limited to 2021. Economic statistics are sourced from the Conference Board of Canada’s 20-year forecast.

3.1 0.3 3 4 DATE: April 28, 2023 Page 31 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary The 2023 Load Forecast Report provides an analysis of electricity demand projections, though specific details are redacted due to confidentiality. The report is part of a regulatory proceeding and includes data relevant to forecasting methodologies and energy usage patterns.

3 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary The 2023 Load Forecast Report provides an analysis of projected electricity demand, incorporating factors such as Heating Degree Days (HDD) and Cooling Degree Days (CDD). The report is part of a regulatory proceeding and includes redacted confidential information.

Page 45 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary The document is a redacted version of the 2023 Load Forecast Report, which likely contains information related to electricity demand forecasting for Nova Scotia. Due to redaction, specific details are not available.

2030 22 13 13 8 7 2031 23 14 14 9 7 2032 24 14 14 10 8 2033 26 15 14 11 8 DATE: April 28, 2023 Page 51 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED 1 4.5 Price Data 2 3 Price data is an input to the...

AI summary The document discusses the methodology for calculating price data used in load forecasts, including the use of a 12-month moving average of real revenue per kWh. It also references a 6.9% annual increase in electricity prices for 2023-2024, as outlined in Schedule 'B' of the GRA Settlement Agreement, followed by an average 2% annual increase thereafter.

a flat profile in real terms. Figure 37 15 shows price forecasts by class. 16 17 Figure 37: Historical and projected real electricity prices (real dollars per kWh) 18 19 20 Settlement Agreement – NS Power General Rate Application – M10431...

AI summary The text discusses the impact of electricity prices on class sales using a price elasticity model estimated at -0.15, provided by Itron. This model is used to forecast load based on historical data and projected price changes.

fferent components for 13 2020, 2021 and 2022 actuals vs forecast and weather normalized totals, and the 2023 14 forecast. 15 16 Figure 39: Comparison of Forecast to Actuals 17 Year 2020 2021 2022 2023 Forecast Sales 4540 4718 4715 4830 We...

AI summary The document presents a comparison of forecasted and actual electricity sales for the years 2020 to 2023, highlighting the impact of weather variance, non-weather variance, and the ongoing influence of the COVID-19 pandemic on residential load forecasts, including assumptions about continued hybrid work models.

energy exports are not included. Figure 53 provides a breakdown of the significant 7 variances between forecast and actuals for 2022. 8 9 Figure 53: 2022 Variance to Actual 10 Res Comm Ind Other Losses NSR 2022 Forecast 4,715 3,091 2,542 7...

AI summary The text discusses energy usage variances in 2022, highlighting the impact of weather, unexplained residential load increases, and factors like continued pandemic restrictions and higher-than-expected heat pump installations. It also forecasts an annual increase in NSR from 2023 to 2033, driven by new customers, space heating, and EV adoption, with some offset from solar, DSM, and RTR.

1 control program offering in 2023, providing a total available peak savings of approximately 2 1 MW by early 2024. 3 4 NS Power is also working with E1 on a two-phase pilot project to investigate automatic 5 and manual control of various...

AI summary NS Power is implementing a demand response (DR) program with E1, aiming for 1 MW of peak savings by early 2024. A two-phase pilot project with commercial and industrial customers is underway, targeting 6.8 MW of peak mitigation. Data from these initiatives will be used to refine load forecasts and is expected to impact the 10-year forecast within the sensitivity analysis.

MW 2023 Forecast Peak 2,256 Interruptible -88 Weather (-22.8°C 12hr lag avg) +250 Wind (28.4 km/h daily avg) +26 Weekend -30 Lighting (estimated) -150 Unexplained +203 2023 Feb Peak 2,467 3 4 The nature of this peak is one that falls outsi...

AI summary The text discusses the challenges in forecasting peak demand for 2023, noting that a regression model fails to accurately predict the peak due to extreme conditions such as weather and unexplained factors. The peak is compared to a P90 estimate, and further analysis is planned for 2024. The Load Forecast is noted as statistically accurate but less useful for assessing individual end-use contributions to peak demand.

Page 94 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED

AI summary The document is a redacted version of the 2023 Load Forecast Report, which discusses energy demand projections and related planning considerations for Nova Scotia. Key topics include forecasting methodology, energy usage patterns, and infrastructure planning.

1 11.0 SENSITIVITY ANALYSIS 2 3 The sales and peak forecasts are fundamentally uncertain and depend on many variables, 4 including economics, weather, adoption of distributed generation, electricity rates and 5 DSM. Although each of these...

AI summary The text discusses the uncertainty in sales and peak forecasts due to factors such as economics, weather, distributed generation, electricity rates, and demand-side management (DSM). A P10/P90 probability analysis using Monte Carlo simulation was conducted in 2017 to estimate the probable distribution of future load, with sensitivity bands shown in Figure 74 representing a range of approximately 434-558 GWh over a 10-year period.

s represent actual system totals. 25 DATE: April 28, 2023 Page 95 of 98 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report REDACTED 1 Figure 74: System Energy Sensitivity 2 3 4 Similarly, a P10/P90 scenario was created f...

AI summary The document discusses the creation of a P10/P90 scenario for peak demand, using random sampling of weather and economic drivers. The variation in peak demand is approximately 376-448 MW, and Figure 75 shows the peak forecast with the latest adjustments to the peak end-use model.

,142 1.7% 2,436 -1.8% 138 99.0% 742 11,288 1.4% 2024 4,847 0.4% 3,074 -2.2% 2,414 -0.9% 99 -27.8% 734 11,168 -1.1% 2025 4,846 0.0% 3,074 0.0% 2,445 1.3% 99 0.0% 734 11,199 0.3% 2026 4,901 1.1% 3,098 0.8% 2,489 1.8% 99 -0.3% 741 11,327 1.1%...

AI summary The text presents a table with forecasted values for various metrics across years 2023 to 2033, including demand and load forecasts. The table is part of the 2023 Load Forecast Report Appendix A, which provides detailed data for analysis.

evening January 11 weekday 2022 155 - 2,061 2,216 12.6 -15 -14 evening 2023 146 4 2,105 2,256 1.8 -14 Forecast 2024 147 12 2,111 2,271 0.7 -14 Forecast 2025 148 24 2,119 2,291 0.9 -14 Forecast 2026 156 36 2,148 2,340 2.1 -13 Forecast 2027...

AI summary The text presents a series of load forecast data from 2022 to 2033, including metrics such as peak demand, load factors, and forecasted values. It is part of an appendix from the 2023 Load Forecast Report, which includes confidential information that has been redacted.

2023 Load Forecast Report Appendix B Page 8 of 33 Appendix B – Forecast Model Details Residential Input Variables – XOther Intensities Econ + Reg Struct Water Cook Ref/Frz Wash/ TV Light Misc Other Coeff Total Heat Dry Use Xother Var 2023...

AI summary The document provides details on residential and commercial load forecasting models, including input variables and their projected changes from 2023 to 2033. The residential model includes variables like water heating, cooking, and lighting, while the commercial model uses an SAE average use model and customer forecasts.

Forecast Report Appendix B Page 17 of 33 Appendix B – Forecast Model Details General Service Model Statistics Model Statistics Iterations 13 Adjusted Observations 120 Deg. of Freedom for Error 110 R-Squared 0.919 Adjusted R-Squared 0.912 A...

AI summary The General Service Model Statistics provide details on the forecast model used in the 2023 Load Forecast Report. The model has a high R-squared value of 0.919, indicating a strong fit, and shows a mean absolute percentage error of 2.56%. The general demand class is forecast as gross total sales rather than average use.

23 Load Forecast Report Appendix B Page 22 of 33 Appendix B – Forecast Model Details Small Industrial Model Statistics Model Statistics Iterations 1 Adjusted Observations 120 Deg. of Freedom for Error 106 R-Squared 0.849 Adjusted R-Squared...

AI summary This section of the Load Forecast Report provides statistical details of the Small Industrial Model, including metrics such as R-squared, adjusted R-squared, AIC, BIC, and other diagnostic statistics. It also outlines the specification of the Medium Industrial Model, which includes variables such as monthly binaries and economic indicators related to manufacturing employment.

ast Report Appendix B Page 25 of 33 Appendix B – Forecast Model Details Medium Industrial Model Statistics Model Statistics Iterations 1 Adjusted Observations 204 Deg. of Freedom for Error 190 R-Squared 0.735 Adjusted R-Squared 0.717 AIC 1...

AI summary This section provides statistical details of a medium industrial load forecast model, including metrics such as R-squared, AIC, BIC, and other diagnostic statistics. The model statistics are presented in a table format with values for various parameters, indicating the model's performance and reliability.

2023 Load Forecast Report Appendix B Page 32 of 33 Appendix B – Forecast Model Details Peak Model Statistics Model Statistics Iterations 1 Adjusted Observations 120 Deg. of Freedom for Error 103 R-Squared 0.986 Adjusted R-Squared 0.983 AIC...

AI summary This section presents statistical details of the peak load forecasting model, including metrics such as R-squared, AIC, BIC, and error statistics. The model has a high R-squared value of 0.986, indicating a strong fit, and the Mean Absolute Percentage Error (MAPE) is 2.21%.

2023 Load Forecast Report Appendix B Page 33 of 33 Appendix B – Forecast Model Details Peak Model Fit As seen in the figure below (and in the model statistics above), this approach produces a good fit with historical data. Although it was...

AI summary The document provides details on the peak model fit and forecast comparison and accuracy for the 2023 Load Forecast Report. The peak model is noted to have a good fit with historical data, though it lacks an explicit peak DSM variable due to insignificant parameters. The forecast comparison includes figures on total energy requirement, system peak demand, and firm peak demand.

REDACTED 2023 Load Forecast Report Appendix C Page 3 of 10 Appendix C – Forecast Comparison and Accuracy Figure C3: Firm Peak Demand Figure C4 below provides an overview of the energy forecast accuracy. For these calculations, the load of...

AI summary This appendix discusses the accuracy of energy forecasts by removing the load from major pulp and paper mills, which contribute significant variance. The analysis shows that the mean absolute percent error (MAPE) is less than 2% for a 5-year forecast, but accuracy decreases beyond that period. Firm peak load forecasts show higher error, averaging under 4% for the first 5 years.

REDACTED 2023 Load Forecast Report Appendix C Page 4 of 10 Appendix C – Forecast Comparison and Accuracy in later years. Figure C6 provides the system peak accuracy, with an average of just under 4 percent in the first 5-year period. REDAC...

AI summary This section of the 2023 Load Forecast Report Appendix C discusses the accuracy of system peak forecasts, noting an average error of just under 4 percent over the first 5-year period, as illustrated in Figure C6.

3 Load Forecast Report Appendix C Page 5 of 10 Appendix C – Forecast Comparison and Accuracy

AI summary This section of the Load Forecast Report provides an appendix comparing forecast data with actual outcomes, focusing on the accuracy of the forecasts. It is part of a larger analysis used in regulatory proceedings.

1.2% -0.7% 2021 -1.6% Statistics Lead Time (Years): 1 2 3 4 5 6 7 8 9 10 REDACTED (CONFIDENTIAL INFORMATION REMOVED) REDACTED 2023 Load Forecast Report Appendix C Page 6 of 10 Appendix C – Forecast Comparison and Accuracy NSR less mills Fo...

AI summary The document presents statistical data on forecast accuracy, including lead time in years and metrics such as average percent error and MAPE for various forecast periods. The data shows varying levels of forecast accuracy across different time horizons.

2023 Load Forecast Report Appendix C Page 7 of 10 Appendix C – Forecast Comparison and Accuracy

AI summary This section of the 2023 Load Forecast Report provides a comparison and accuracy analysis of load forecasts, focusing on the data presented in Appendix C, Page 7 of 10.

Figure C5: Firm Peak Forecast Accuracy Firm Peak Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast for: for: for: for: for: for: for: for: for: for: Issued 2013 2014 2015 2016 2017 2018 2019...

AI summary Figure C5 presents a table showing firm peak forecast accuracy from 2012 to 2022, with actual firm peak values provided for 2013, 2014, and 2015. The data compares forecasted values against actual values, illustrating discrepancies over time.

2021 2,062 Actual Firm Peak: 1,897 2,036 1861.3 2013.6 1,951 1,993 1,949 1954 1,875 2061 Percent Error 2012 3.2% -4.0% 4.5% -3.9% -0.9% -3.3% -1.0% -1.2% 2.8% -6.8% 2013 -4.7% 4.8% -3.3% -0.4% -2.9% -0.3% -0.8% 3.2% -6.6% 2014 4.2% -4.1% -...

AI summary This table presents actual firm peak values and percent errors from 2012 to 2021, showing fluctuations in performance over time. The data indicates varying levels of accuracy in forecasting, with some years showing significant errors, such as -10.4% in 2015 and 11.2% in 2020.

11.2% -0.4% 2021 0.1% Statistics Lead Time (Years): 1 2 3 4 5 6 7 8 9 10 REDACTED (CONFIDENTIAL INFORMATION REMOVED) REDACTED 2023 Load Forecast Report Appendix C Page 8 of 10 Appendix C – Forecast Comparison and Accuracy Firm Peak Forecas...

AI summary The document presents statistical data on load forecast accuracy over different time horizons, including average percent error and MAPE values for forecasts from 1 to 10 years ahead. The data shows varying levels of accuracy, with some periods showing positive errors and others showing negative errors.

2023 Load Forecast Report Appendix C Page 9 of 10 Appendix C – Forecast Comparison and Accuracy

AI summary This section of the 2023 Load Forecast Report provides a comparison and accuracy analysis of load forecasts. It is part of Appendix C and appears on page 9 of 10.

Figure C6: System Peak Forecast Accuracy System Peak Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast Forecast for: for: for: for: for: for: for: for: for: for: Issued 2013 2014 2015 2016 2017 2018...

AI summary This table presents the system peak forecast accuracy over various years, comparing forecasts issued in different years with actual system peak values. The data shows discrepancies between forecasts and actual values, highlighting forecasting challenges in the electricity sector.

13.6% -0.2% 2021 0.4% Statistics Lead Time (Years): 1 2 3 4 5 6 7 8 9 10 REDACTED (CONFIDENTIAL INFORMATION REMOVED) REDACTED 2023 Load Forecast Report Appendix C Page 10 of 10 Appendix C – Forecast Comparison and Accuracy System Peak Fore...

AI summary This section discusses the accuracy of load forecasts over a 10-year period, with statistics showing average percent error and MAPE values. It also outlines the use of Monte Carlo simulation in a sensitivity analysis for economic and weather variables.

Forecast Sensitivity Analysis Sensitivity Analysis The P10/P90 sensitivity analysis used Monte Carlo simulation for the economic and weather variables. The algorithm uses the following sequence 1. Once the deterministic SAE class regressio...

AI summary A sensitivity analysis using Monte Carlo simulation is performed to assess how variations in weather and economic factors impact the SAE class regression models. Historical data from the past 20 years is used to model these variations as normal distributions.

D 2023 Load Forecast Report Appendix D Page 3 of 9 Appendix D – Forecast Sensitivity Analysis Figure D1: Distribution of January HDD 5. Oracle’s Crystal Ball runs about 10,000 trials, taking a random set of numbers from the relevant variab...

AI summary The document discusses the use of Oracle’s Crystal Ball for running 10,000 trials in a Monte Carlo simulation to analyze load forecast sensitivity. It highlights the inclusion of heat pumps in the SAE models and the distribution of energy and peak load forecasts before the impact of demand-side management (DSM).

before the impact of DSM). 10th (10%) and 90th (90%) percentiles can easily be obtained from Normal distributions and so they are highlighted in D2. Figure D2: Distribution of Energy (Before DSM) Page 3 of 8 REDACTED (CONFIDENTIAL INFORMAT...

AI summary The text discusses probabilistic load forecasting, focusing on the distribution of energy and peak demand before demand-side management (DSM) is applied. It references figures showing percentiles and sensitivity analysis, including the impact of variables on system peak forecasts.

of 8 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Appendix E Page 1 of 21 MARCH 31, 2023 2023 10yr Preliminary Forecast . REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Appendix E Page 2 of 2...

AI summary This document outlines the timeline and purpose of the 2023 10-year preliminary load forecast report, including stakeholder engagement and submission to the NSUARB. It provides an overview of the forecast updates and improvements, and sets the stage for feedback and discussion.

age 18 of 21 Forecast Comparison – Peak The peak forecast is similar to the 2022 forecast, and slightly higher by 2032 due to higher EV penetration and increased new customer count. 18 . REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Loa...

AI summary The peak forecast remains similar to the 2022 forecast but is expected to increase slightly by 2032 due to higher EV penetration and more new customers. The 2022 forecast showed variances compared to actuals, including impacts from weather, residential usage, and wind generation.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-1: 2 3 Reference: Exhibit N-1, p. 22, lines 13-15. 4 5 Has NS Power revised its operational dispatch planning...

AI summary NS Power explains that its operational dispatch load forecast model uses an artificial neural network with weather variables (temperature, wind speed, cloud cover) and historical load data, achieving 2.5% MAPE accuracy. It clarifies that this model differs from long-term planning models but does not incorporate lagged average temperature or wind speed adjustments.

cast? 25 (i) If so, please provide the supporting analysis. 26 (ii) If not, please identify how NS Power will support additional market 27 interventions. 1 Itron, Cold Climate Heat Pump Load Study (July 2022), pp. 17-18. Submitted as Appen...

AI summary The text requests analysis on supporting market interventions and references an Itron study submitted with NS Power's On-Bill Financing Final Report (M09321). It also cites the 2023 Load Forecast Report (NSUARB M11108) as part of the regulatory proceeding dated June 20, 2023.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) details NSPI's responses to information requests from the Consumer Advocate. The report addresses energy usage projections and related regulatory considerations.

2 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL Date/time Month Day Time System Total EV Available Load Power (kW) Vehicles [MW] 1/11/2022 18:00 1 11 18 2215.7 44.57 84...

AI summary NSPI submitted a non-confidential response to the Consumer Advocate's information requests regarding the 2023 Load Forecast Report (NSUARB M11108). The report includes time-stamped data on system load, total EV power, and available vehicles, reflecting energy demand patterns and EV integration impacts.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL Date/Time (hour Peak (MW) Weekday? Peak Period? ending) 2022-01-11 18:00 2216 Yes Yes 2022-01-11 19:00 2179 Yes Yes 2022-01...

AI summary NSPI provides load forecast data and discusses demand reduction from TVP rates in response to the Consumer Advocate's information requests, referencing the 2023 Load Forecast Report (NSUARB M11108). The data includes peak load measurements and mentions demand reduction outcomes dependent on the TVP pilot's conclusion.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-6: 2 3 Reference: Exhibit N-1, Figure 73, p. 94. 4 5 Please confirm that the AMI Weather Normalized values use...

AI summary NSPI confirms in its response to the Consumer Advocate that the AMI Weather Normalized values in the 2023 Load Forecast Report use the 2023 Load Forecast method (12-hr temperature lag and wind speed). The response is part of NSUARB M11108.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-7: 2 3 Reference: Exhibit N-1, Figures 15-17; Exhibit N-1, M10569, Figures 15-17. 4 5 Relative to the 2022 Loa...

AI summary The Consumer Advocate requested explanations for changes in the 2023 Load Forecast Report, including decreased household compensation forecasts, population and residential customer projections, and historical data revisions. NSPI referenced prior responses (IR-4 part c) and provided a figure comparing 2022 and 2023 forecasts for residential customers and population changes.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 2 3 The changes compared to the 2022 forecast include increased population growth and a 4 corresponding increase in housi...

AI summary NSPI's 2023 Load Forecast Report (NSUARB M11108) cites increased population growth and housing completions as key factors compared to the 2022 forecast. Responses to the Consumer Advocate's information requests direct reference to NSUARB IR-4 sections (b)-(e).

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-9: 2 3 Reference: Exhibit N-8, M10569, pp. 4-5; Exhibit N-12, M10569; Exhibit N-1, Attachment 5 4 Residential...

AI summary The Consumer Advocate questions NSPI's 2023 Load Forecast Report regarding residential model accuracy, specifically Heating Degree Day (HDD) trends and model changes since 2022. Requests include confirming NSPI's response to prior critiques and detailing model modifications. The matter is tied to NSUARB proceeding M11108.

trend, as referenced in part (d). 26 27 (f) Please state describe any changes made by NS Power to its small general load model 28 related to any changes described in part (e). 29 Date Filed: June 20, 2023 NSPI (CA) IR-9 Page 1 of 3 2023 Lo...

AI summary NSPI responds to the Consumer Advocate's request regarding changes to its small general load model, clarifying that the WtXHeat variable includes factors beyond HDD. Removing the declining HDD trend shows an increase in WtXHeat, contradicting Mr. Wilson's initial observation about the variable's decline.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 that the effect of the decreasing number of HDD is almost perfectly offset by additional 2 heat pumps (with a number of o...

AI summary NSPI explains that decreasing heating degree days (HDD) in residential classes are offset by increased heat pump adoption, while small general classes see greater offsets from electrification, aligning with 2022 forecast trends. The analysis emphasizes the correct interpretation of WtXHeat behavior in load forecasting.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) details NSPI's responses to information requests from the Consumer Advocate. The report addresses energy usage projections and related regulatory considerations.

3). 27 28 (c) Please confirm that the example calculation of the XOther value (p. 7) represents the 29 model formula (p. 2). For each of the calculation components below, if not confirmed, Date Filed: June 20, 2023 NSPI (CA) IR-10 Page 1 o...

AI summary The Consumer Advocate requests confirmation that the XOther value calculation in the 2023 Load Forecast Report aligns with the model formula, as part of NSPI's response to information requests under NSUARB proceeding M11108.

Where HDDy,m is the Heating Degree Day 1 for a given month m of the year y, 0 F 28 HHSize is the average household size, ResEcon is Employment Compensation 1 Throughout the report we use HDD18 which is the largest between 0 and 18-daily av...

AI summary The text discusses the use of Heating Degree Day (HDD18) calculations in load forecasting, referencing NSPI's 2023 Load Forecast Report (NSUARB M11108) and responses to the Consumer Advocate's information requests. HDD18 is defined as the maximum of 0 and 18 minus daily average temperature, added monthly.

Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 1 divided by House Hold population, Price is the price of electricity for the specific 2 customer class. Each variable uses its...

AI summary NSPI responds to the Consumer Advocate's information requests regarding load forecasting methodologies, confirming certain calculations and providing formulas for variables like CoolUse, which involves Cooling Degree Days, household data, and price factors. The text explains regression-based elasticity calculations and clarifies unconfirmed claims.

𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅15 𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 15 12 13 Where CDDy,m is the Cooling Degree Day 2 for a given month m of the year y, 1 F 14 HHSize is the average household size, ResEcon is Employment Compensation 15 divided by House Hold population, Price is...

AI summary The document discusses load forecasting methodology, including variables like Cooling Degree Days (CDD), household size, and electricity pricing. It references regression coefficients (e.g., b2) and appliance usage multipliers from Attachment 1, responding to a Consumer Advocate inquiry on NSPI's 2023 Load Forecast Report (NSUARB M11108).

Load Forecast Report (NSUARB M11108) NSPI Responses to Consumer Advocate Information Requests NON-CONFIDENTIAL 𝑆𝑆𝑆𝑆𝑆𝑆𝑦𝑦𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 � � � 𝐸𝐸𝐸𝐸𝐸𝐸𝑦𝑦𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 1 𝑂𝑂𝑂𝑂ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑦𝑦 ,𝑚𝑚 = � 𝐸𝐸𝐸𝐸15 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 × 𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 × 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑚𝑚𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 𝑆𝑆𝑆𝑆...

AI summary The Load Forecast Report (NSUARB M11108) details NSPI's responses to the Consumer Advocate's information requests. The document includes non-confidential technical formulas involving variables like 𝑆𝑆𝑆𝑆𝑆𝑆𝑦𝑦𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇 and 𝐸𝐸𝐸𝐸𝐸𝐸𝑦𝑦𝑇𝑇𝑇𝑇𝑇𝑇𝑇𝑇, potentially related to forecasting methodologies or energy usage calculations.

s given 22 in the table following the quoted text, it is associated to WtXOther. 23 24 (d) Not confirmed. Please refer to the responses for parts (a), (b), and (c) above. 25 Date Filed: June 20, 2023 NSPI (CA) IR-10 Page 4 of 5 2023 Load F...

AI summary NSPI responds to the Consumer Advocate's information requests regarding the 2023 Load Forecast Report, explaining categorization of 'Dish' under XOther via OtherIndex and referencing Attachment 1 for formulas in the Load Forecast Report.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to EfficiencyOne Information Requests NON-CONFIDENTIAL 1 Request IR-1: 2 3 Reference: NS Power 2023 Load Forecast, Page 33, Lines 6-7 4 “The estimated number customers who installed...

AI summary NSPI's response to EfficiencyOne's information request explains that the 20,800 heat pump installations in 2022 were calculated using monthly contractor surveys. The estimate is based on data collected through these surveys, as detailed in Attachment 1.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to EfficiencyOne Information Requests NON-CONFIDENTIAL 1 Request IR-3: 2 3 Reference: NS Power 2023 Load Forecast, Page 36, Figure 23 4 Please provide a version of Figure 23 that inc...

AI summary EfficiencyOne requested a complete version of Figure 23 from NSPI's 2023 Load Forecast Report covering 2023-2033. NSPI directed them to Synapse IR-7 Attachment 1, specifying data locations in tabs/columns, noting E3 data is only available in 5-year increments.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-2: 2 3 In section 4.2 Weather Data on page 17 of the Application, the Heating Degree Day (HDD) 4 uses a reference tempera...

AI summary NSPI explains that HDD18 is standard in Canadian utilities and aligns with Environment Canada and US EIA. It is a key component of the SAE model, with provincial examples showing HDD18's correlation to residential sales in 2019 and 10-year averages.

NSPI (NSUARB) IR-2 Page 1 of 3 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 2 3 4 5 The figures above highlight two issues while modeling monthly residential sales as a linear 6...

AI summary NSPI highlights two issues in modeling residential sales using HDD and SAE: sensitivity loss near the y-axis for low HDD values and improved model fit with higher reference temperatures. It also advises reducing regression complexity by limiting HDD/CDD variables to avoid overlap.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-3: 2 3 In section 4.3 Economic Information on page 27, lines 16 to 17 of the Application state: 4 “Housing completion num...

AI summary NSPI responded to NSUARB's query about adjusting the 2023 load forecast using housing data, stating no outliers were identified in the forecast period and confirming no separate forecast exists for houses under construction.

1 Request IR-4: 2 3 In reference to Figures 17 to 19 on pages 29 to 31 of the Application: 4 5 (a) Page 28 of the Application notes that financial variables have been adjusted to 6 constant dollars to eliminate inflation effects. Have all...

AI summary The document contains questions about data adjustments, verification with CMHC, and revisions in load forecasts related to residential and commercial economic drivers, seeking explanations for discrepancies and impacts on forecasts.

1 (i) The figures for Non-Manufacturing Employment from 2013 onward have 2 been revised from the 2022 Load Forecast report. Notably, the 2022 forecast 3 provided an employment high of 452,000 occurring in 2031 and 2032, whereas 4 this fore...

AI summary The text presents questions regarding revised employment and GDP data in the 2023 load forecast, querying the reasons for changes, their impact on the forecast, and validation against other forecasts.

ring employment with positive manufacturing GDP 24 growth. 25 (iv) Has the data been checked against other forecasts? 26 27 Response IR-4: 28 29 (a) Yes, indicators in dollars are adjusted for the historic period. 30 Date Filed: June 20, 2...

AI summary NSPI confirmed that data in the 2023 Load Forecast Report (NSUARB M11108) was adjusted for historical periods to align with other forecasts. The response addresses validation of load forecasting methodologies and data consistency.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 (b) New Construction data is provided by the Conference Board of Canada, who list the 2 CMHC as one of their sources. NS Power is no...

AI summary NSPI explains that changes in historical data from the Conference Board of Canada and Statistics Canada do not affect load forecasts, as forecasts use indexed values and apply data only in the forecast period. NSPI lacks alternative data sources for household compensation and non-manufacturing employment trends.

rcent. 28 29 Revised average hourly wages were 2.6 percent higher than unrevised wages 30 over the period from January 2006 to December 2022. After the initial upward Date Filed: June 20, 2023 NSPI (NSUARB) IR-4 Page 4 of 6 2023 Load Forec...

AI summary The text discusses a 2.6% increase in revised average hourly wages compared to unrevised wages from 2006 to 2022, as part of NSPI's response to NSUARB's information requests regarding the 2023 Load Forecast Report.

1 revision to wage estimates in 2006, trends in year-over-year wage growth 2 among employees were equivalent on average in the revised and unrevised 3 wage series. In December 2022, revised average hourly wages grew 4.8% 4 (+$1.49 to $32.6...

AI summary The text discusses revised wage estimates in 2006, noting similar year-over-year growth trends between revised and unrevised series. It highlights a 4.8% wage increase in December 2022, linking this to household compensation and employment forecasts. The analysis compares GDP and employment forecasts from Canadian banks with those of the Conference Board of Canada, noting discrepancies in 2023-2024 projections.

The bank forecasts were from November 2022 through January 2023, with most 22 dating from December 2022, so it was felt that the more recent Conference Board 23 forecast (from February 2023) would be more appropriate and therefore no 24 ad...

AI summary NSPI responds to NSUARB information requests regarding the 2023 Load Forecast Report, noting that Conference Board of Canada (CBoC) data on GDP and employment was used without explanations for changes in GDP or negative growth. NSPI highlights discrepancies in impacts on small vs. medium industrial forecasts.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-5: 2 3 Page 33 of the Application lines 5 to 6 state: “The end-use model uses an estimated saturation 4 of 40 percent of...

AI summary NSPI explains the increase in heat pump saturation from 33% to 40% in the 2023 Load Forecast Report, citing higher-than-forecast installations in 2022. The response references prior forecasts and installation data to justify the adjustment.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-7: 2 3 Page 38 of the Application, lines 12 to 13 state: “It is estimated that there were approximately 4 1900 EVs in the...

AI summary NSPI confirmed its 2022 EV estimate of 1900 vehicles was based on Nova Scotia government open data (updated to 1922 EVs as of January 2023). The load forecast model combined historical EV adoption data with federal targets (20% in 2026, 60% in 2030, 100% in 2035) to project growth rates.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-9: 2 3 Page 53 of the Application lines 1 to 2 state “The SAE models are estimated using a -0.15 4 price elasticity.” 5 6...

AI summary NSPI responds to NSUARB's query about the SAE model's -0.15 price elasticity in the 2023 Load Forecast Report. The elasticity is short-run, applied to consumption variables, and forecast to remain constant despite electrification. All rate classes use the same elasticity, with no model results indicating a need for revision.

ting. To put the impact of 30 the price elasticity into perspective, the following table shows the estimated impact to sales 31 in 2033 for various estimates of price elasticity: Date Filed: June 20, 2023 NSPI (NSUARB) IR-10 Page 1 of 2 20...

AI summary The text analyzes the impact of price elasticity on residential electricity sales forecasts for 2033, showing varying sales changes (-1 to -4 GWh) based on elasticity estimates. It notes that while price elasticity affects sales, its impact is minor compared to other variables in the regression model, with the SAE model equations referenced in CA IR-10.

12 to estimate their monthly contributions, as this is the only way to shape the annual 30 amounts into monthly amounts. In the case of July values of WtXHeat being higher that Date Filed: June 20, 2023 NSPI (NSUARB) IR-11 Page 1 of 2 2023...

AI summary The document references NSPI's responses to NSUARB information requests regarding the 2023 Load Forecast Report, highlighting the need for monthly contribution estimation to annualize data. The filing date is June 20, 2023, with the matter labeled as NSUARB M11108.

1 WtXCool, it is due to the imperfect translation of annual values to monthly amounts 2 combined with the fact that the heating variable is a factor of 10 higher than the cooling 3 variable, so any residual amounts may be larger in heating...

AI summary The text explains that the SAE model improves monthly load forecasts by using monthly weather data and billing information, but annual-to-month translation imperfections create larger residuals in heating. Lagged weights allocate heating/cooling usage across three months, with the model relying on weighted sums of prior and current monthly data for accuracy.

1 Request IR-13: 2 3 Figure 42: Residential Sales Components by Year on page 60 of the Application project 54 4 GWh for New Customers in 2023 and 106 GWh for 2024. In the 2022 Load Forecast the 5 New Customer forecast was 56 GWh for 2022 a...

AI summary The request challenges the overstatement of new customer growth forecasts in residential sales, citing discrepancies between annual and cumulative figures. The response clarifies that the figures are cumulative, with incremental data provided to justify the forecasted growth.

3 Forecast GWh (incremental) 54 52 2023 Forecast New Customers (incremental) 6057 6023 17 Date Filed: June 20, 2023 NSPI (NSUARB) IR-13 Page 1 of 1 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON...

AI summary NSPI responds to NSUARB's IR-14 by citing E3's EV forecast data, which allocates 35% of energy and 30% of peak to commercial class. This categorization includes medium/heavy-duty vehicles under commercial, while residential charging is segmented by location.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-15: 2 3 Page 63 of the Application presents the forecast for Commercial Class sales and identifies 4 that sales lag the r...

AI summary NSPI explains that no specific consideration was given to an economic downturn due to unpredictability, relying instead on economic variables from the Conference Board’s forecast to account for broader economic changes in load forecasting models.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-16: 2 3 Page 63 of the Application indicates that a COVID variable was added to the General rate 4 class model to account...

AI summary NSPI explains a COVID variable in its load forecast model, formulated as a pseudo-binary with decreasing values from 2020 to 2033, applied to adjust commercial load estimates. The model's statistical significance will be re-evaluated annually. NSPI acknowledges potential changes in commercial sales-GDP/employment relationships but states no specific analysis was conducted for this forecast.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-17: 2 3 Page 65 of the Application states in line 13 “the addition of EV loads adds 272 GWh by 4 2023….” Please explain h...

AI summary NSPI clarifies that the 272 GWh EV load addition referenced in the 2023 Load Forecast Report applies to 2033, not 2023, and is based on aggregated provincial data from E3. The response notes that more detailed EV usage data will be incorporated as it becomes available.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-18: 2 3 Page 67 of the Application indicates that Large General Service growth is driven by several 4 hospital projects....

AI summary NSPI responds to NSUARB's IR-18 request, stating that load growth from provincial health authority projects will begin in 2024, peaking in 2025-2027. The long-term forecast accounts for sustained consumption but excludes specific DSM investments like solar panels. NSPI lacks confidence in 2023 project completion.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-19: 2 3 Page 68 of the Application states that Small Industrial sales “have been flat for the last 10 4 years and are exp...

AI summary NSPI responds to NSUARB's IR-19 and IR-20 requests regarding the 2023 Load Forecast Report. For Small Industrial sales, GDP is used directly in the model with a regression coefficient of 2.156. Medium Industrial sales declines are attributed to RTR load loss, not employment changes.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-21: 2 3 Figure 51: New Large Industrial Projects forecast. 4 5 (a) What is the confidence in this forecast level of new p...

AI summary NSPI responds to NSUARB's request regarding confidence in the 2023 Load Forecast Report's new industrial project energy requirements. Confidence decreases over time, starting at 65% in 2023 and dropping to 10% by 2027. No high or low estimates were provided, and DSM is applied at the class level, not specific projects.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL 1 Request IR-23: 2 3 Page 74 of the Application states “From 2023 to 2033, NSR is forecast to increase at an 4 average of 0.7 percent...

AI summary NSPI explains that the increase in NSR forecast from 2023 to 2033 is due to higher customer growth, increased space heating, and a higher EV sales forecast incorporating federal mandates, rather than new factors being added to the model.

NSPI (NSUARB) IR-26 Page 1 of 2 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to NSUARB Information Requests NON-CONFIDENTIAL Number of Customer Year Upgrade Orders 2019A 1698 2020A 1860 2021A 2502 2022A 2937 2023F 3113 1 2 (c)...

AI summary NSPI provided responses to NSUARB information requests regarding the 2023 Load Forecast Report. The report includes customer upgrade orders from 2019 to 2023 and notes that the sensitivity analysis only considers weather and economics, while the peak forecast accounts for electrification demand impacts but not customer panel size changes.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Small Business Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-1: 2 3 Please reference Synapse IR-3 parts A-C, and reference Section 4.0, p 16. 4 5 (a) Please provide...

AI summary NSPI explains that the 10-year regression period (2013–2023) was chosen for most classes due to statistical significance, except Medium Industrial, which uses an econometric model with manufacturing employment data. They note that pre-2013 data may not reflect recent demographic/technological changes, but a 20-year forecast with only 10 years of data risks poor model fit.

6 that produce statistically insignificant results or models that indicate a loss of relevance to 27 the underlying input data, alternate historic time periods may be explored. 28 Date Filed: June 20, 2023 NSPI (SBA) IR-1 Page 1 of 2 2023...

AI summary NSPI defends its 2023 load forecast methodology, stating a 10-year timeframe was chosen for sufficient data points and trend analysis. It confirms no changes to 2022 assumptions, nor benchmarking against alternative datasets or testing alternate assumptions.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Small Business Advocate Information Requests NON-CONFIDENTIAL 1 Request IR-7: 2 3 Please reference Figure 51: New Large Industrial Projects. 4 5 (a) Has NS Power performed any sen...

AI summary NSPI confirmed it did not perform sensitivity or historical analysis on the actualization of new large industrial projects, as requested by the Small Business Advocate in relation to the 2023 Load Forecast Report (NSUARB M11108).

ment 01 Residen�al Intensi�es, tabs Intensi�es, Shares, HP and StructuralVars Figure 26,30 Synapse IR-09 Atachment 1 Figure 27 Synapse IR-44 Atachment 2 Figure 28 Synapse IR-09 Atachment 3 Figure 31 Synapse IR-10 Atachment 1 Figure 32 Syna...

AI summary The document references the 2023 Load Forecast Report (NSUARB M11108) and NSPI's responses to Synapse Energy Economics' information requests. It lists figures from the 2022 and 2023 LFR, along with Synapse attachments related to intensity, pricing, and forecast classes.

l General Figure 47 2023 LFR Atachment 04 10 year Forecast Classes, tab General Demand Figure 49 2023 LFR Atachment 04 10 year Forecast Classes, tab Small Industrial Figure 50 2023 LFR Atachment 04 10 year Forecast Classes, tab Medium Indu...

AI summary The document outlines the submission of the 2023 Load Forecast Report (LFR) by NSPI, including attachments and figures related to demand forecasting across various sectors. It references Synapse Energy Economics' information requests and associated regulatory filings under NSUARB proceeding M11108.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-2: 2 3 Historical Sales and Energy Data (Figures 1, 3, 39-40, 44, 46-50, 52) 4 5 (a) Please provide in...

AI summary NSPI provided responses to Synapse Energy Economics' information requests related to the 2023 Load Forecast Report (NSUARB M11108), including historical sales data, usage data, system load data, and unmetered sales details. Data was provided in electronic attachments, with specific references to spreadsheet tabs and columns.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-3: 2 3 Billed versus Accrued Sales and Regression Period (Section 4.0, p 16) 4 5 (a) The report notes t...

AI summary NSPI responds to Synapse Energy Economics' queries about the 2023 Load Forecast Report's methodology, addressing billed vs. accrued sales differences, regression periods for industrial forecasts, and statistical impacts of timeframe adjustments. References to Attachment 1 and page 28 highlight concerns about model relevance when using shorter timeframes for industrial data.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-3 Attachment 1 Page 1 of 1 Accrued Sales 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Residential 4,394 4,370 4,484 4,318 4,374 4,581 4,664 4,652 4,661 4...

AI summary The 2023 Load Forecast Report (LFR) by Synapse presents historical sales data (2013-2022) for residential, small/general, and industrial customer categories, distinguishing between accrued and billed sales metrics. The data shows trends in energy consumption across sectors, with residential and medium industrial categories showing the highest growth.

249 253 257 247 245 254 261 253 257 264 Medium Industrial 492 471 476 463 463 472 461 468 476 483 Residential variance -0.7% 0.8% 0.4% -1.3% -0.2% -0.8% 0.5% 0.5% -0.9% 0.6% Small General variance -0.1% 0.7% 0.1% -1.1% -1.0% -0.2% 0.0% -0....

AI summary The document includes load forecast data with variances across residential, industrial, and commercial sectors, referencing the 2023 Load Forecast Report (NSUARB M11108) and NSPI's responses to Synapse Energy Economics' information requests. Non-confidential data highlights percentage variations in demand forecasts.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-4: 2 3 Weather Data (Section 4.2, pp 17-26) 4 5 (a) Please provide in electronic spreadsheet format the...

AI summary NSPI responds to Synapse Energy Economics' information requests regarding the 2023 Load Forecast Report (NSUARB M11108), seeking weather data, HDD/CDD calculations, peak load data, and methodology for forecasting. Requests focus on data transparency and verification of forecasting techniques.

Absolute Percentage Error (MAPE)? Date Filed: June 20, 2023 NSPI (Synapse) IR-4 Page 1 of 2 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Reques...

AI summary NSPI responds to Synapse Energy Economics' inquiry about the use of MAPE in the 2023 Load Forecast Report. NSPI explains that MAPE was selected as a common and appropriate accuracy measure, based on regression models for single/multiple weather station methods. No other approaches were investigated.

e better predictive properties it 23 would result in a lower MAPE. No other approaches were investigated. MAPE is a 24 common and appropriate measure of forecast accuracy. 1 Historical Data - Climate - Environment and Climate Change Canada...

AI summary The document discusses the use of a forecasting method with better predictive properties to reduce MAPE (Mean Absolute Percentage Error) in the 2023 Load Forecast Report. NSPI's responses to Synapse Energy Economics' information requests are referenced, with a focus on forecast accuracy metrics.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-5: 2 3 Economic Information (Section 4.3, pp 26-31) 4 5 (a) Please provide in electronic spreadsheet fo...

AI summary The document outlines Synapse Energy Economics' information requests to NSPI regarding the 2023 Load Forecast Report (LFR). Requests focus on data sources, model construction (e.g., work-from-home variables, housing completions), population trends, GDP/employment drivers for commercial/industrial models, and labor/automation considerations in industrial forecasts.

other drivers considered? Why were they not 27 chosen? 28 29 (g) What consideration if any was given to labour shortages or increased automation for 30 the industrial sector? Date Filed: June 20, 2023 NSPI (Synapse) IR-5 Page 1 of 4 REDACT...

AI summary The document contains questions from a regulatory proceeding regarding the 2023 Load Forecast Report (LFR), including considerations for unchosen drivers and labor/automation impacts on industrial sectors. NSPI is responding to Synapse Energy Economics' information requests under NSUARB matter M11108.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 2 (h) What would be the effects on the industrial forecast if 10 years were used for the 3 regression? 4 5 (i) Please provide the inflation adjustments used to convert to constant dollars. 6 7 (j) Regarding economic forecasts, identify t...

AI summary The response addresses questions about economic forecasts, regression models, and data sources. It references Conference Board of Canada data, major banks' comparisons, and pandemic-related variables affecting load. The new housing forecast is sourced from the Conference Board, with definitions linked to CMHC.

fessionals/housing-markets-data-and- Date Filed: June 20, 2023 NSPI (Synapse) IR-5 Page 2 of 4 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Req...

AI summary The document references the 2023 Load Forecast Report (NSUARB M11108) and NSPI's responses to Synapse Energy Economics' information requests. It is part of a regulatory proceeding involving load forecasting and data disclosure, with non-confidential content filed on June 20, 2023.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 research/housing-research/surveys/methods/methodologies-starts-completions-market- 2 absorption-survey. 3 4 d) Population increased in 2022 due mainly to immigration from other provinces and from 5 outside the country (immigration repres...

AI summary Population growth in Nova Scotia, driven by immigration, is projected to increase from 2021 to 2033, leading to higher housing completions and customer counts. Non-manufacturing GDP and employment data are used in forecasts to model these trends.

538,157 0.5% 11 12 e) The non-manufacturing GDP and non-manufacturing employment variables have been 13 used in previous forecasts and provided consistent models with good model fit statistics. 14 Unless there is reason to believe they are...

AI summary The text discusses variable selection in load forecasting models, noting that non-manufacturing GDP and employment variables provided consistent results in prior forecasts. For industrial models, GDP and manufacturing employment were selected due to better model statistics, while other variables like manufacturing GDP and exports performed worse.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 g) Labour shortages and increased automation were not considered explicitly, but any 2 significant factors that would impact the employment or GDP variables would be included 3 in the underlying forecast provided by the Conference Board...

AI summary The text discusses forecasting methodologies for energy demand, noting that labor shortages and automation impacts are implicitly considered via the Conference Board of Canada's (CBoC) GDP forecasts. Using a 10-year period for the Medium Industrial model reduces statistical relevance of employment variables, with adjusted R-squared dropping from 0.72 to 0.37. Forecasts use annual CPI for inflation adjustments and rely on CBoC's 'most likely future' economic scenarios, with discrepancies checked against bank forecasts in Attachment 1.

of Canada’s view of the most likely future. 15 16 k) Bank forecasts are checked for discrepancies in the near-term GDP and employment 17 forecasts, and are provided in Attachment 1. Date Filed: June 20, 2023 NSPI (Synapse) IR-5 Page 4 of 4...

AI summary The text references the review of bank forecasts for discrepancies in near-term GDP and employment projections, with details provided in Attachment 1. It also cites the 2023 Load Forecast Report (Synapse IR-5) as part of the documentation.

2023 Load Forecast Report Synapse IR-5 Attachment 1 Page 1 of 3

AI summary This document is an attachment from the 2023 Load Forecast Report by Synapse, part of a regulatory proceeding in Nova Scotia. It likely outlines energy demand projections and methodologies used for forecasting, relevant to NSPI and the NSUARB's oversight of utility planning.

CBoC 20 Year Forecast (Dec 2022) CBoC 5 Year Forecast (Feb 2023) + 20 year forecast Calculated (Average (Average (Average Aggregation) (Average (Average (Average Aggregation) Aggregation) Current (Average Aggregation) (Average Aggregation)...

AI summary The text references the Conference Board of Canada's (CBoC) 20-year and 5-year economic forecasts, including housing completions, real GDP at basic prices, employment data, and the Consumer Price Index (CPI). These forecasts are used for regulatory analysis in Nova Scotia.

ll Prices by Industry, Total, Nova Nova Scotia, Consumer Price Singles, Nova Multiples, Nova Compensation of Prices by Industry, All Prices by Industry, Total, Nova Nova Scotia, Consumer Price Scotia Scotia of employees; Industries, Nova M...

AI summary The text presents a table with economic data on prices by industry in Nova Scotia, including the Consumer Price Index, compensation of employees, GDP, and employment statistics, using indices and monetary values from 2006 onwards.

41946.0 46386.7 3630.7 513.0 34.5 2.0 2898 42756 479 21381 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-5 Attachment 1 Page 2 of 3 Bank Source GDP Emp Housing CPI 2023 2024 2023 2024 2023 2024 2023 2024...

AI summary The document contains economic data from multiple banks (BMO, RBC, TD, National Bank, Scotiabank) projecting Nova Scotia's GDP, employment, housing, and CPI for 2023-2024, contextualized within the 2023 Load Forecast Report (Synapse IR-5 Attachment 1). The data reflects varying economic outlooks and serves as input for energy demand forecasting.

IAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-5 Attachment 1 Page 3 of 3 CPI (CBoC 2Household Household Household Median emMedian emMedian employment income (indexed 2012) 2012 1.250833 20910 16716.86 1 28100 22465.02 1 201...

AI summary The text presents economic data from the 2023 Load Forecast Report (LFR) by Synapse, including CPI (CBoC), household income, employment metrics indexed to 2012. This data is used for load forecasting in regulatory proceedings, reflecting trends in income, employment, and inflation to model energy demand and cost factors.

1 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 Household Income Median Income REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CO...

AI summary The text references the 2023 Load Forecast Report (NSUARB M11108) and NSPI's responses to Synapse Energy Economics' information requests. It includes redacted income data tables and non-confidential sections related to regulatory proceedings involving load forecasting and utility responses.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-6: 2 3 Residential End-Use Intensity Trends (Section 4.4, pp 32-51) 4 5 (a) Please provide in electroni...

AI summary NSPI provided data from NRCan and U.S. EIA for residential and commercial models in response to Synapse Energy Economics' IR-6 request, detailing adjustments made to align intensities with NRCan reports and NS Power billing data. Attachments reference specific data tables and modifications in the 2023 Load Forecast Report.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) is referenced alongside NSPI's responses to Synapse Energy Economics' information requests. The document pertains to forecasting methodologies and regulatory processes involving NSPI and Synapse.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

End Use Consumption Data End-Use Stock Data Lighting Table 3 Assumed 100% of households have lighting Room and central air- Table 4 Table 27 conditioning Electric furnace and heat Table 9 Table 21 pumps Electric hot water Table 10 Table 28...

AI summary The document details end-use consumption data compilation methods, referencing tables from Attachment 2 and external sources like EIA or surveys. It specifies which tables populate the 'NRCanDetail' tab in the 2023 LFR Attachments 2 and 3 for various end uses, including space heating, electric hot water, and lighting.

Table 32 Outdoor Lighting Table 34 Total Consumption Table 1 Date Filed: June 20, 2023 NSPI (Synapse) IR-6 Page 2 of 3 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy E...

AI summary NSPI explains its methodology for aligning NRCan data with residential billing records in the 2023 Load Forecast Report (LFR), using a scaling factor derived from calibration comparisons to adjust annual end-use intensity values.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Single Detached Energy Use (PJ) 32.6 34.3 35.2 32.9 31.2 29.8 30.3 34.7 35.0 35.8 35.7 38.2 33.9 33.0 31.4 32.4 29.1 29.7 31.3 31.4 E...

AI summary The table presents total single detached energy use and energy use by end-use in Nova Scotia from 2000 to 2019. It shows energy use in petajoules (PJ) for categories such as space heating, water heating, appliances, lighting, and space cooling over the years.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Single Attached Energy Use (PJ) 2.7 2.8 2.9 2.7 2.6 2.5 2.5 2.9 3.0 3.1 3.1 3.4 3.1 3.0 2.9 3.0 2.8 2.8 3.0 3.1 Energy Use by Energy...

AI summary This table presents the total single attached energy use and energy use by energy source in Nova Scotia from 2000 to 2019. Electricity and heating oil are the primary energy sources, with heating oil showing a decline over time, while electricity use remains relatively stable.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Energy Use for Wholesale Trade (PJ) 3.3 3.3 3.3 3.4 4.0 3.6 3.2 3.7 3.5 3.1 2.8 3.5 3.4 3.1 3.4 3.4 3.3 3.1 3.0 3.0 Energy Use by Ene...

AI summary The table presents historical data on total energy use for wholesale trade in Nova Scotia from 2000 to 2019, broken down by energy source. It shows fluctuations in energy consumption across electricity, natural gas, fuel oils, and other sources over the years.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Energy Use for Retail Trade (PJ) 8.4 8.8 8.7 9.2 10.9 10.6 9.5 11.0 10.4 9.2 8.4 10.3 10.3 9.4 10.2 10.5 10.1 9.6 9.2 9.3 Energy Use...

AI summary The text provides a table showing total energy use for retail trade in Nova Scotia from 2000 to 2019, broken down by energy source. It includes data for electricity, natural gas, light fuel oil, heavy fuel oil, steam, and other energy sources, measured in petajoules (PJ).

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Energy Use for Health Care and Social Assistance (PJ) 8.5 8.6 8.4 8.5 10.2 9.2 8.3 9.3 8.7 7.7 6.8 8.6 8.8 8.3 9.3 9.5 9.2 8.7 8.5 8....

AI summary The table presents total energy use and energy use by source for the Health Care and Social Assistance sector from 2000 to 2019, measured in petajoules (PJ). Electricity, natural gas, light fuel oil, and heavy fuel oil are the primary energy sources tracked over the period.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Total Auxiliary Equipment Energy Use (PJ) 7.1 7.8 7.6 8.9 9.2 9.3 9.3 9.9 10.0 10.5 10.8 10.5 9.9 9.9 10.7 10.4 10.8 9.4 10.1 10.7 Energy U...

AI summary The table presents the total auxiliary equipment energy use and energy use by energy source, measured in petajoules (PJ), from 2000 to 2019. Electricity is the primary energy source, with natural gas, fuel oil, and other sources showing no usage.

Water Heating Energy Use for Offices1 (PJ) 0.5 0.5 0.5 0.7 0.5 0.5 0.5 0.6 0.5 0.5 0.6 0.7 0.6 0.5 0.5 0.5 0.4 0.4 0.4 0.4 Energy Use by Energy Source (PJ) Electricity 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0...

AI summary The text presents data on energy use for water heating in offices, measured in petajoules (PJ), across various energy sources including light fuel oil, natural gas, and other sources. The data appears to be tabular and may be part of a larger analysis or report on energy consumption patterns.

Water Heating Energy Use for Educational Services (PJ) 0.6 0.5 0.5 0.6 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.7 0.7 0.5 0.6 0.6 0.5 0.5 0.5 0.5 Energy Use by Energy Source (PJ) Electricity 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.0...

AI summary The document presents data on water heating energy use for educational services in PJ units, categorized by energy sources such as electricity, natural gas, fuel oil, and others. The data spans multiple years, showing variations in energy consumption.

Water Heating Energy Use for Accommodation and Food Services (PJ) 0.3 0.3 0.3 0.3 0.4 0.3 0.3 0.4 0.3 0.3 0.3 0.4 0.4 0.3 0.3 0.3 0.3 0.3 0.3 0.3 Energy Use by Energy Source (PJ) Electricity 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0...

AI summary The text presents energy use data for water heating in accommodation and food services, detailing the use of various energy sources such as light fuel oil, kerosene, natural gas, and others, measured in petajoules (PJ) across multiple time periods.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Space Cooling Energy Use for Other Services (PJ) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.1 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Energy Use...

AI summary The text presents a table showing energy use data for space cooling in 'Other Services' from 2000 to 2019. It details energy use by source (electricity and natural gas) and their respective shares, along with floor space activity in million square meters over the same period.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 2 (h) Please document the data that the RESHAPE model used for this forecast and its 3 impacts on the forecast. 4 5 Response IR-7: 6 7 (a) Please see the following table estimating the number of customers for each category: 8 Customers C...

AI summary The response provides a table detailing the estimated number of customers in various categories, including residential, electric resistance, and heat pump users, from 2023 to 2029. This data is part of the RESHAPE model's forecast and its impact analysis.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

t all- 22 electric and all-electric and their space heating energy intensity (kWh/m2). Combined 23 with floor area estimates for a typical small and medium/large commercial customer 24 derived from CBECS, E3 estimated the space heating dem...

AI summary The document discusses the estimation of space heating energy intensity for commercial customers in Nova Scotia, using data from CBECS and NS Power, and scaling a sample of buildings from the New England region to represent the province's commercial heating service demand fuel mix. Load forecast data for 2025 and 2030 are also presented.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) is referenced alongside NSPI's responses to Synapse Energy Economics' information requests. The document pertains to forecasting methodologies and regulatory processes involving NSPI and Synapse.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

ide the confidential version of the January 2023 Smart Grid Nova Scotia 28 Semi-Annual Report as an attachment. 29 30 (h) Please provide the source data and calculations for Figure 28. Date Filed: June 20, 2023 NSPI (Synapse) IR-9 Page 1 o...

AI summary The document includes requests for data and calculations related to the 2023 Load Forecast Report, specifically Figure 28 and Figure 30. NSPI provides responses referencing attachments and outlines the methodology for applying federal sales targets to annual sales forecasts.

(2) No funded project between Jan 2020 and Dec 2020? Guidance and Definitions Innovation Valid 1) Between Jan 2020 and Dec 2020,has your business or institution developed any new or (2) No significantly improved products as a result of the...

AI summary The text contains a series of questions regarding the development of new or significantly improved products and services between January 2020 and December 2020, with responses indicating 'No'. It also includes references to redacted confidential information and attachments from reports related to load forecasting and smart grid initiatives.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 Request IR-16: 2 3 Demand Side Management (Section 4.6, pp 54-56). 4 5 (a) Please provide the source data for the DSM values used in this forecast. 6 (b) Please provide the DSM values used in the latest IRP. 7 (c) Please provide the DSM...

AI summary The request seeks source data for DSM values used in the forecast, including the latest IRP and E1 potential study. The response indicates that DSM values for 2023-2025 are based on EOne’s Settlement Plan and values beyond 2025 are based on the EOne Potential Study. A table is referenced with energy and demand values for various years.

29.6 26.5 0.425 0.397 2027 71.6 55.7 9.8 41.2 39.5 30.4 26.0 0.425 0.397 2028 73 62.9 11.1 42.0 44.6 31.0 29.4 0.425 0.397 2029 73.7 52.2 9.2 42.4 37.0 31.3 24.4 0.425 0.397 2030 73.3 51 9 42.1 36.2 31.2 23.8 0.425 0.397 2031 74.1 48 8.5 4...

AI summary The text provides a table with numerical data and references a 2023 Load Forecast Report (NSUARB M11108) and NSPI responses to Synapse Energy Economics information requests. The data appears to relate to load forecasting and energy planning.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-17: 2 3 Demand Side Management Adjustment (Section 4.6, pp 54-56) 4 5 (a) Please provide the details of...

AI summary The document outlines a request for information regarding the Demand Side Management (DSM) adjustment in the 2023 Load Forecast Report. It asks for details on the data and statistical analysis used to develop the DSM coefficient for residential and commercial/industrial sectors, as well as any changes compared to the 2022 report and statistical measures associated with the coefficients.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-17 Attachment 1 Page 1 of 9

AI summary The document is a redacted version of the 2023 Load Forecast Report, specifically Attachment 1, which is part of Synapse IR-17. The content has been removed due to confidentiality.

0.00 2019 1 373,071.51 57,266.26 0.00 112,180.64 142,159.89 38,806.00 0.00 0.00 0.00 0.00 0.00 2019 2 352,964.42 55,508.64 0.00 113,712.53 135,331.90 38,813.00 0.00 0.00 0.00 0.00 0.00 2019 3 355,810.95 53,744.36 0.00 108,482.34 132,702.32...

AI summary The text includes a table with numerical data for the year 2019, followed by a reference to a redacted section of the 2023 Load Forecast Report by Synapse, Attachment 1, Page 2 of 9.

2.55 128,630.43 39,609.00 0.00 0.00 0.00 0.00 1.00 2025 4 68,827.42 0 98,247.50 135,081.59 39,609.00 0.00 0.00 0.00 0.00 1.00 2025 5 63,208.91 536.4 68,480.38 134,215.99 39,609.00 0.00 0.00 0.00 0.00 1.00 2025 6 56,560.29 5,275.16 41,497.5...

AI summary The text contains a table with numerical data and a reference to the 2023 Load Forecast Report Synapse IR-17 Attachment 1 Page 3 of 9, which has been redacted due to confidentiality.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-17 Attachment 1 Page 3 of 9

AI summary The text is a redacted page from the 2023 Load Forecast Report, specifically Attachment 1, page 3 of 9. No substantive content is visible due to redaction, so no summary of key arguments or facts can be provided.

REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-17 Attachment 1 Page 4 of 9

AI summary The text is a redacted portion of the 2023 Load Forecast Report, specifically Attachment 1, Page 4 of 9, from Synapse IR-17. It contains confidential information that has been removed.

.36 130,023.47 39,609.00 0.00 0.00 0.00 0.00 1.00 2033 10 66,533.17 14,081.58 27,236.99 129,220.29 39,609.00 0.00 0.00 0.00 0.00 1.00 2033 11 72,729.97 293.69 63,811.87 130,283.99 39,609.00 0.00 0.00 0.00 0.00 1.00 2033 12 76,546.78 0 104,...

AI summary The text contains a table with numerical data and a reference to the 2023 Load Forecast Report, specifically Synapse IR-17 Attachment 1, Page 5 of 9. The content is partially redacted, indicating the presence of confidential information.

49.601 51,174.019 123,531.380 150,601.931 0.000 0.000 0.000 2017 6 296,797.412 -13,110.703 2,535.060 32,161.561 124,566.433 150,645.062 0.000 0.000 0.000 2017 7 287,339.111 -12,696.831 13,158.080 12,185.247 123,851.502 150,841.112 0.000 0....

AI summary The text contains a table of numerical data spanning multiple months and years, likely related to financial or operational metrics. It includes a reference to the 2023 Load Forecast Report, specifically Synapse IR-17 Attachment 1, Page 6 of 9, with a note that certain information has been redacted due to confidentiality.

7 0.000 0.000 0.000 2022 9 300,718.111 -23,516.960 42,648.144 2,983.357 123,265.064 155,338.507 0.000 0.000 0.000 2022 10 313,826.220 -26,048.462 7,911.523 14,741.485 122,540.146 155,220.876 0.000 39,460.651 0.000 2022 11 285,290.550 -28,6...

AI summary The text provides numerical data spanning multiple months and years, likely related to financial or operational metrics. It includes values such as load forecasts, costs, and other figures, with the mention of the Load Forecast Report (LFR) and Synapse IR-17 Attachment 1. The data appears to be part of a larger analysis or report.

080.99 -28,846.48 0 118,112.92 109,507.42 155,307.139 0.000 0.000 0.000 2033 4 348,486.32 -27,298.64 0 105,454.60 115,023.22 155,307.139 0.000 0.000 0.000 2033 5 318,384.66 -25,070.20 317.262 73,519.43 114,311.03 155,307.139 0.000 0.000 0....

AI summary The text presents a series of numerical entries, likely related to financial data or load forecasts, with a mention of the 2023 Load Forecast Report, Synapse IR-17, Attachment 1, Page 9 of 9, which has been redacted due to confidentiality.

0.000 0.000 0.000 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-17 Attachment 1 Page 9 of 9 Variable Coefficient StdErr T-Stat P-Value MSales.EESavingsProfiled -0.397 0.129 -3.076 0.27% MStructGen.WtXCool...

AI summary The document presents statistical data from the 2023 Load Forecast Report, including coefficients and p-values for various factors affecting load forecasting. It is part of NSPI's responses to Synapse Energy Economics' information requests in the NSUARB M11108 proceeding.

loads would be expected to decline. Date Filed: June 20, 2023 NSPI (Synapse) IR-18 Page 2 of 3 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Req...

AI summary The document refers to the 2023 Load Forecast Report (NSUARB M11108) and provides details on how estimated floor space is calculated based on historic data and forecasted additions for residential properties.

2,035,725) (4,037,108) REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-18 Attachment 1 Page 5 of 6

AI summary The text includes a redacted section from the 2023 Load Forecast Report, specifically Synapse IR-18 Attachment 1, Page 5 of 6. The content appears to be part of a regulatory proceeding in Nova Scotia, likely related to energy forecasting or load management.

t the EV load, sales would decrease 23 due to increased solar, RTR sales, and DSM. Please refer to Appendix B pages 12 and 18 24 for a breakdown of the components of the forecast. Date Filed: June 20, 2023 NSPI (Synapse) IR-20 Page 1 of 1...

AI summary The 2023 Load Forecast Report indicates that the addition of EV load to the Small General Service class has significantly increased load growth, contributing 21% to the class load and 3.9% overall growth in 2023 compared to 1.5% in 2022. The increase is primarily driven by the heating component and is expected to continue through 2027.

tor to the increase starting in 2027. Date Filed: June 20, 2023 NSPI (Synapse) IR-21 Page 1 of 1 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information R...

AI summary The document is a non-confidential response from NSPI to information requests by Synapse Energy Economics regarding the 2023 Load Forecast Report (NSUARB M11108), filed on June 20, 2023.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-22: 2 3 General Service (Section 6.2). 4 5 (a) Please explain and quantify the specific reasons for the...

AI summary The 2023 Load Forecast Report (NSUARB M11108) outlines responses from NSPI to Synapse Energy Economics' information requests. Key factors affecting load changes include EV load, RTR participation, DSM programs, and increased efficiency. The report notes EV load added 11.4% to class load, while RTR reduced load by 2.7%. DSM program effects decreased slightly compared to the 2022 forecast.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

Res Modeled C&I Large Firm Inter. System Heat EV DR DSM Peak Elect. Cust. Peak Cust. Peak Peak (MW) (MW) (MW) (MW) (MW) (MW) (MW) (MW) (MW) (MW) 2022 2,011 2 3 -4 2 107 -14 2105 146 2,255 2023 2,028 3 6 -12 3 109 -26 2111 147 2,271 2024 2,...

AI summary The table presents modeled peak demand and related factors for various years, including heat, electric vehicle (EV), demand response (DR), and demand-side management (DSM) contributions. The text explains the concept of 'unexplained' components in forecasting models, which account for uncertainty and differences between predictions and actual outcomes.

cannot explain. All forecasting models have uncertainty in their predictions, and the 4 unexplained item attempts to quantify this. It is calculated as the difference between 5 forecast minus actuals. 6 7 (h) Please refer to Attachment 2....

AI summary The document discusses the 2023 Load Forecast Report, which includes an efficiency study for Nova Scotia from 2021 to 2045. It references an earlier study prepared by Navigant for EfficiencyOne and filed in August 2019. The report addresses forecasting models and their uncertainties.

Cumulative Savings as a Percent of Total Sales (%) Source: Navigant analysis ©2019 Navigant Consulting, Ltd. Page 7 . REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-30 Attachment 1 Page 16 of 355 Nova Scot...

AI summary The document presents cumulative savings as a percentage of total sales and discusses the technical and economic potential of energy efficiency and demand response in Nova Scotia over a 25-year period, showing flat savings potential of 2,100 to 2,200 MW (gross at generator) and market potential ranging from 375 MW to 600 MW (net at generator).

Load (%) Source: Navigant analysis ©2019 Navigant Consulting, Ltd. Page 10 . REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-30 Attachment 1 Page 19 of 355 Nova Scotia Energy Efficiency and Demand Response...

AI summary The document discusses the market winter peak demand savings potential from energy efficiency (EE) and demand response (DR) in Nova Scotia, estimating scenarios ranging from 15% to 24% over a 25-year period, as analyzed by Navigant Consulting.

than the base scenario costs due to lower participation levels and lower per participant incentives and marketing costs. The annual portfolio costs (in nominal dollars) are expected to increase from: • $3.3 million in 2021 to $21.4 million...

AI summary The document outlines projected annual portfolio costs for demand response (DR) scenarios from 2021 to 2045, showing increasing costs across base, high, and low scenarios. It also highlights the achievable MW potential and percent of peak load reduction for cost-effective DR options under each scenario, with the base scenario representing the highest potential.

Study for 2021-2045 potential results are presented for DR options, sub-options, customer class, and building type for cost- effective DR options. Section 12 – presents the Conclusion of the study. The report also includes the following el...

AI summary The document outlines a study on energy efficiency and demand response potential from 2021 to 2045, including appendices with modeling plans, baseline studies, and model inputs and outputs for residential and commercial sectors.

Electronics & IT Other Source: Navigant 2.3 Fuel Shares Navigant developed fuel share and equipment data for each end use based on the segmentations defined in the previous sections, using the 2019 Nova Scotia Baseline Study results for sp...

AI summary The document outlines the methodology used by Navigant to develop fuel share and end use allocation data based on the 2019 Nova Scotia Baseline Study and the 2018 End Use Intensity Model. This data is used to understand energy consumption patterns and is essential for forecasting and planning purposes.

used to develop the reference case forecasts. Figure 2. Data Collection Granularity for Base Year / Reference Forecast Input Parameters Source

AI summary The text discusses the use of input parameters sourced for developing reference case forecasts, with a mention of data collection granularity in the base year and reference forecast, as illustrated in Figure 2.

he 26 accuracy of the model? 27 Date Filed: June 20, 2023 NSPI (Synapse) IR-31 Page 1 of 5 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Request...

AI summary The document is a response from NSPI to information requests regarding the 2023 Load Forecast Report, filed under the NSUARB proceeding M11108. It includes a question about the accuracy of the model used in the report.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

experiment is to compare top-down (current) and bottom-up class peak estimation 10 (experiment), therefore, to compare apples to apples DSM can be removed from both 11 models. 12 Date Filed: June 20, 2023 NSPI (Synapse) IR-31 Page 4 of 5 R...

AI summary The text discusses a comparison between top-down and bottom-up methods for estimating class peak demand, suggesting that removing DSM from both models allows for a more accurate comparison. It also references the 2023 Load Forecast Report and mentions the expectation of having sufficient AMI data for rate classes by the 2024 load forecast report.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-32: 2 3 Sensitivity Analysis (Section 11 and Appendix D) 4 5 (a) Please provide in electronic format th...

AI summary NSPI provided responses to Synapse Energy Economics' information requests regarding the 2023 Load Forecast Report. The response includes details on the sensitivity analysis, the selection of weather and economic variables, and the statistical distributions used in the Monte Carlo simulation.

Attachment 2 for the inputs used. 30 Date Filed: June 20, 2023 NSPI (Synapse) IR-32 Page 1 of 2 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Re...

AI summary The document discusses challenges in modeling DSM and end uses due to limited data sets and difficulties with the probabilistic approach in the 2023 Load Forecast Report. These issues are under study for inclusion in future models.

IR-32 Page 2 of 2 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-32 Attachment 1 Page 1 of 2 Year Actual_Sales p10 p50 p90 2013 11,193.6 2014 11,037.4 2015 11,099.1 2016 10,809.0 2017 10,872.7 2018 11,248....

AI summary This document provides a table showing historical and projected electricity sales data from 2013 to 2033, including actual sales and forecast percentiles (p10, p50, p90) for each year. The data is part of the 2023 Load Forecast Report from Synapse.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 Request IR-33: 2 3 Sensitivity Analysis (Section 11, 2020 IRP Comparison, pp 97-98) 4 5 (a) Please provide information about how the evergreen IRP analysis might affect future 6 loads. 7 8 (b) For comparison with the previous forecast pl...

AI summary The response to Request IR-33 discusses the impact of the Evergreen IRP analysis on future loads and provides a comparison of load forecasts for 2023 and 2032. The Evergreen IRP results will help NS Power update its IRP Action Plan and inform future load forecasts, particularly as electrification and peak reduction programs progress.

2,106 Electrification 20 Date Filed: June 20, 2023 NSPI (Synapse) IR-33 Page 1 of 1 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests PARTI...

AI summary The document refers to the 2023 Load Forecast Report (NSUARB M11108) and includes NSPI's responses to information requests from Synapse Energy Economics. It is marked as partially confidential and redacted.

1 Request IR-34: 2 3 Appendix A: Forecast 4 5 (a) Please provide in electronic format the specific calculations used to create the values 6 in Tables A1, and A2. 7 8 (b) Please explain the historical changes in the Interruptible Contributi...

AI summary The request asks for the calculations behind Tables A1 and A2 and an explanation of changes in the Interruptible Contribution to Peak values. The response directs to the 2023 Load Forecast Report and provides a table showing historical and forecasted interruptible peak values, noting an expected slow increase from 146-156 MW.

centage for Other. 26 27 (h) Has NSPI collaborated with E1 in developing common assumptions for the end-use 28 components? If so, please describe the results of that collaboration. 29 Date Filed: June 20, 2023 NSPI (Synapse) IR-35 Page 1 o...

AI summary The document contains a portion of NSPI's responses to information requests related to the 2023 Load Forecast Report, specifically addressing collaboration with E1 on common assumptions for end-use components.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Response IR-35: 2 3 (a) Please refer to 2023 Load Forecast Report Attachment 5 - Residential Model, filed 4 electr...

AI summary NSPI provides responses to Synapse Energy Economics' information requests regarding the 2023 Load Forecast Report. Key details include heat and cool share forecasts for residential use, references to specific report attachments, and a note on no collaboration in 2022.

Patterns and the economic indicator for that forecast report (Employment Compensation 27 being this year’s). 28 29 (h) No collaboration on this topic was undertaken in 2022. Date Filed: June 20, 2023 NSPI (Synapse) IR-35 Page 2 of 2 REDACT...

AI summary The text mentions the 2023 Load Forecast Report and notes that no collaboration on the topic occurred in 2022. It includes a filing date and a reference to a confidential document.

9 2,249.83 607.97 61.95 1.01 29.78 260.86 0.00 58.87 1.21 2031 1,335.35 2,356.85 621.76 61.11 1.01 30.32 270.40 0.00 59.23 1.22 2032 1,260.74 2,459.29 635.04 60.20 1.01 30.85 279.41 0.00 59.65 1.23 2033 1,190.05 2,552.41 647.75 59.18 1.00...

AI summary The text presents a table with numerical data and a reference to the 2023 Load Forecast Report Synapse IR-35 Attachment 1 Page 2 of 8, which has been redacted due to confidentiality.

354.78 379.60 0.00 1,186.38 0.97 2028 1,709.65 529.47 374.75 45.88 185.08 51.06 48.98 761.21 343.01 376.75 0.00 1,187.15 0.97 2029 1,737.98 529.42 370.21 45.51 183.19 50.93 49.21 761.69 331.54 373.92 0.00 1,186.90 0.97 2030 1,752.69 529.36...

AI summary The text presents numerical data spanning multiple years, likely related to financial or operational metrics. It includes values such as costs, revenues, and other figures, but the content is partially redacted. The mention of the '2023 Load Forecast Report' suggests the data may be related to energy forecasting or planning.

.01 0.00 0.00 1.00 1.00 1.00 1.00 0.00 0.00 2031 1,401.01 0.00 0.00 1.00 1.00 1.00 1.00 0.00 0.00 2032 1,401.01 0.00 0.00 1.00 1.00 1.00 1.00 0.00 0.00 2033 1,401.01 0.00 0.00 1.00 1.00 1.00 1.00 0.00 0.00 REDACTED (CONFIDENTIAL INFORMATIO...

AI summary The text appears to be a portion of a Load Forecast Report from 2023, specifically Attachment 1, Page 4 of 8. The content is partially redacted, indicating that it contains confidential information. The document includes numerical data spanning from 2031 to 2033, likely related to energy load forecasts.

ENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-35 Attachment 1 Page 4 of 8

AI summary The document refers to the 2023 Load Forecast Report by Synapse, specifically Attachment 1, which is part of a regulatory proceeding. It provides context for forecasting methodologies and energy usage patterns.

349.5 5,620.7 4,392.0 422.6 5,454.3 1401.00923 0 0 1 1 1 1 0 0 0 0 307.59972 4,594.02 490.69 5,121.63 2031 4,375.1 360.0 5,621.2 4,418.7 439.8 5,455.8 1401.00923 0 0 1 1 1 1 0 0 0 0 307.59972 4,622.00 510.61 5,123.01 2032 4,415.3 369.9 5,6...

AI summary The text presents a table with numerical data and references to a 2023 Load Forecast Report (Synapse IR-35 Attachment 1 Page 6 of 8), which has been redacted due to confidentiality. The table includes figures related to load forecasts and other metrics.

333 (290) 29.3 5,366.2 2031 10,208.5 488,654.0 950.8 13,871.8 2,492.9 29,574.5 532,100.3 16000 4860 1.037016 379.2 5,390.6 (14.0) 433 (348) 70.9 5,461.5 2032 10,297.5 488,654.0 845.1 14,716.9 2,313.3 31,887.8 535,258.8 16000 4860 1.040964...

AI summary The document presents data on residential load forecasts, including customer numbers, EVs, solar installations, and demand-side management (DSM) impacts from 2023 to 2033. It highlights changes in these metrics over the decade, with notable increases in customer numbers and solar installations, and decreases in residential sales due to DSM efforts.

182 57 1.1 1.161 343 2033 31 288 60 1.2 1.161 549 Change to 1.6% 30.9% 0.9% 11.8% 0.0% 60.1% Xother Inputs EWHeat ECook Ref/Frz Washer/Dr TV Light Misc OtherUse VCoeff Total Xother 2023 1,536 530 645 808 407 504 1,236 0.98 0.939 5,198 2033...

AI summary The document presents data from the 2023 Load Forecast Report (NSUARB M11108), including energy usage projections and changes between 2023 and 2033. It includes energy consumption data for various categories and percentage changes, as well as NSPI's responses to Synapse Energy Economics information requests.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-36: 2 3 Appendix B: Small General Service Model (pp 9-14) 4 5 (a) Please provide in electronic spreadsh...

AI summary NSPI provided responses to Synapse Energy Economics' information requests regarding the 2023 Load Forecast Report. The responses include references to attachments and methodology used in forecasting software, such as Metrix ND, for calculating variables like XHeat, XCool, and XOther.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 is outlined in Appendix B, Forecast Model Details in the Load Forecast report. The inputs 2 to XHeat, XCool, and X...

AI summary The 2023 Load Forecast Report (NSUARB M11108) outlines the methodology and data inputs used by NSPI in their responses to Synapse Energy Economics information requests. The report includes attachments detailing model components, heat pump growth, and assumptions for PV, EV, and DSM forecasts.

TIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-36 Attachment 1 Page 1 of 11

AI summary The text references the 2023 Load Forecast Report, specifically Synapse IR-36 Attachment 1, which is part of a regulatory proceeding in Nova Scotia. The document appears to be a technical attachment related to load forecasting and may be used in a regulatory process.

35,888.06 1.400 2026 64,913.84 1.270 35,772.68 1.430 2027 65,800.60 1.280 35,659.92 1.460 2028 67,031.99 1.290 35,457.38 1.490 2029 68,185.27 1.290 35,350.68 1.520 2030 68,838.53 1.300 35,246.40 1.550 2031 69,852.87 1.300 35,144.50 1.580 2...

AI summary The text presents numerical data related to load forecasts and associated rates for various years from 2026 to 2033. It includes values and percentages, likely representing projections for energy usage and costs. The document is part of a Load Forecast Report and contains confidential information that has been redacted.

DENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-36 Attachment 1 Page 4 of 11

AI summary The text refers to the 2023 Load Forecast Report, specifically Synapse IR-36 Attachment 1, which is on page 4 of 11. The content is not fully visible due to the redaction of sensitive information.

0.00 2032 1.00 1.00 1.00 0.00 0.00 0.00 0.00 2033 1.00 1.00 1.00 0.00 0.00 0.00 0.00 REDACTED (CONFIDENTIAL INFORMATION REMOVED) 2023 Load Forecast Report Synapse IR-36 Attachment 1 Page 5 of 11

AI summary The document contains a redacted section from the 2023 Load Forecast Report, specifically Attachment 1, Page 5 of 11. It includes numerical data and is part of a regulatory proceeding in Nova Scotia.

9,577.1 2029 68,185.3 35,350.7 128,543.5 6,157.1 1,880.7 9,505.8 2030 68,838.5 35,246.4 125,917.3 6,264.3 1,912.1 9,393.4 2031 69,852.9 35,144.5 123,414.4 6,356.6 1,943.5 9,293.1 2032 70,794.3 35,133.4 121,297.7 6,541.4 1,979.8 9,242.9 203...

AI summary The text includes numerical data related to load forecasting, with years ranging from 2029 to 2033 and values for various metrics. It also references a confidential section of the 2023 Load Forecast Report, specifically Synapse IR-36 Attachment 1 Page 7 of 11, and includes terms such as 'ARMA', 'XHeatNew', and 'XCoolNew'.

0.00 0.00 0.00 2.25 0.00 2025 0.00 0.00 0.00 0.00 0.00 2.25 0.00 2026 0.00 0.00 0.00 0.00 0.00 2.25 0.00 2027 0.00 0.00 0.00 0.00 0.00 2.25 0.00 2028 0.00 0.00 0.00 0.00 0.00 2.25 0.00 2029 0.00 0.00 0.00 0.00 0.00 2.25 0.00 2030 0.00 0.00...

AI summary The text presents a table with numerical data spanning years from 2025 to 2033, likely related to energy load forecasts or financial figures. A mention of 'REDACTED (CONFIDENTIAL INFORMATION REMOVED)' and a reference to the '2023 Load Forecast Report Synapse IR-37 Attachment 1 Page 4 of 9' indicate the content is part of a regulatory proceeding involving energy forecasting.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL 1 Request IR-38: 2 3 Appendix B: Small Industrial Model (pp 21-23) 4 5 (a) Please provide in electronic spreadsheet...

AI summary NSPI provided responses to Synapse Energy Economics' information requests regarding the 2023 Load Forecast Report. They explained that the Small Industrial Model uses Metrix ND software and that ManGDP was selected as the long-term economic driver due to its satisfactory predictive performance.

Year Month ResEndUse.ResCool NResEndUse.SmlGenCool NResEndUse.GenCool mVars.CoolLoad mVars.Days mVars.Cool_AvgMW mPkDayWthr.PkCDDIdx mVars.Cool_Var 2030 2 - - - - 28.0 - - - 2030 3 - - - - 31.0 - - - 2030 4 6.9 0.5 3.8 11.1 30.0 0.0 - - 20...

AI summary The text presents a table with data spanning from 2030 to 2031, containing various metrics related to cooling load, days, average megawatts, and peak cooling degree day indices. The data appears to be related to energy usage and load forecasting, with specific values for different months and years.

- - 2033 2 - - - - 28.0 - - - 2033 3 - - - - 31.0 - - - 2033 4 7.7 0.5 4.0 12.2 30.0 0.0 - - 2033 5 2,767.9 181.2 1,437.3 4,386.3 31.0 5.9 - - 2033 6 22,036.6 1,444.9 11,454.1 34,935.6 30.0 48.5 0.2 9.8 REDACTED (CONFIDENTIAL INFORMATION R...

AI summary The text contains a table with numerical data and a reference to a redacted confidential document titled '2023 Load Forecast Report Synapse IR-41 Attachment 1 Page 13 of 19'. The table likely represents load forecasting data, but the content is partially redacted.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 Request IR-45: 2 3 General Forecast Report Improvements – The Board Decision of October 31, 2022 notes that 4 NS Power agreed to the following changes in the 2023 forecast report (p 4): 5 6 In its Reply, NS Power addressed the concerns r...

AI summary NS Power has agreed to improve its 2023 forecast report by incorporating multi-hour temperature and windspeed analysis, multi-station weather data, and electrification impacts. It will also refine DR estimates, evaluate heat pump impacts, and update EV adoption rates, among other changes.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 NS Power’s reply submission did not agree with a few of the intervenors’ 2 requests. NS Power does not consider the line loss determination model as a 3 tool for load planning and said a report on this topic is not required. NS Power 4 c...

AI summary NS Power disagrees with some intervenors' requests, including the need for a line loss determination model report and an ELCC factor for the LIIR. However, it agrees to consider an ELCC adjustment for EV load shapes. The response includes incorporating agreed changes into the report, such as analyzing multi-hour temperature and windspeed impacts on peak load forecasting and incorporating water heating load control with a peak sensitivity analysis.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

1 Request IR-46: 2 3 General Forecast Report Improvements – The Board Decision of October 31, 2022 Includes 4 the following findings (pp 5-6): 5 6 The Board notes that NS Power has agreed to many of the intervenors’ 7 recommendations and a...

AI summary The Board directed NS Power to implement intervenors' recommendations in the Load Forecast report, including detailed analysis of electrification impacts, new technologies, and weather-normalized calculations. It also encouraged continued stakeholder engagement and early sharing of forecast evaluations.

forecast early in the 28 process. 29 30 In addition to the above directives, the Board encourages NS Power to include 31 information on each of the following in future load forecasts: 32 33 • Examine the elasticity used in the SAE model to...

AI summary The NSUARB encourages NS Power to improve future load forecasts by examining the elasticity used in the SAE model and evaluating input variables in the residential model to enhance model robustness.

2023 Load Forecast Report (NSUARB M11108) NSPI Responses to Synapse Energy Economics Information Requests NON-CONFIDENTIAL

AI summary The 2023 Load Forecast Report (NSUARB M11108) includes NSPI's responses to information requests from Synapse Energy Economics. The document is marked as non-confidential and relates to regulatory proceedings involving load forecasting and data disclosure.

and peak from electrification as well as from new technologies, especially findings Please refer to NSUARB Please refer to NSUARB IR- from pilot projects. IR-45. 45. NSUARB NS Power is directed to continue Direction evaluating and implemen...

AI summary The NSUARB has directed NS Power to continue evaluating and implementing improvements to the calculation of weather-normalized values for energy and peak, as recommended by intervenors. This is part of the 2023 Load Forecast Report (NSUARB M11108), which also includes an assessment of the impact of DSM, Solar PV, EVs, and battery storage.

gory UARB Comments Status Notes assess the impact of DSM, Solar PV, EVs, battery storage, and weather. The Board directs NS Power to continue NS Power held a to actively engage with intervenors as stakeholder conference on well as other st...

AI summary The NSUARB has directed NS Power to engage stakeholders and assess the impact of DSM, Solar PV, EVs, and battery storage on load forecasting. The Board also requested an evaluation of elasticity inputs in the SAE model and residential model robustness.

- Given the continued population growth in Nova Scotia and ongoing housing shortage in the province, re-evaluate the use of housing completions for the near term; NSUARB - Given the current inflationary Suggestions environment, evaluate th...

AI summary The document suggests re-evaluating the use of housing completions as a metric for demand forecasting in light of population growth and housing shortages. It also proposes using median household income instead of total household income to better reflect economic conditions, and suggests incorporating demographic factors for more accurate load forecasting.

achieved. Not included forecast. Revisit the short-term economic inputs These metrics are updated provided by the Conference Board of Included. annually and included in Date Filed: June 20, 2023 NSPI (Synapse) IR-46 Page 3 of 4 REDACTED (C...

AI summary The document discusses the 2023 Load Forecast Report and NSPI's responses to Synapse Energy Economics' information requests. It highlights the need to revisit economic inputs from the Conference Board of Canada and EV adoption rates in alignment with Statistics Canada data. NS Power is encouraged to maintain communication with customers regarding large infrastructure projects.

Evidence Regarding Nova Scotia Power’s 2023 Load Forecast Evidence RE: M11108 Prepared for the Nova Scotia Utility and Review Board July 18, 2023 AUTHORS David White, PhD Ben Havumaki Selma Sharaf 485 Massachusetts Avenue, Suite 3 Cambridg...

AI summary This document provides evidence regarding Nova Scotia Power’s 2023 load forecast, prepared for the Nova Scotia Utility and Review Board. It includes an introduction outlining forecast comparisons, sector and DSM (demand-side management) overviews, and recommendations from a prior forecast review. Authors include David White, Ben Havumaki, and Selma Sharaf.

..................................................................................4 1.4. Recommendations from the Previous Forecast Review .....................................................6 2. ENERGY FORECAST .............................

AI summary The document outlines a regulatory proceeding focusing on energy and peak demand forecasting, including DSM effects, sensitivity analysis, and responses to prior recommendations. Sections cover residential, commercial, and industrial sectors, electrification trends, and methodological approaches to forecasting.

) adjustments applied to the SAE forecast values. Third, NSPI applies other factors to reflect customer growth and specific program adjustments. We will discuss all these components in this evidence. 1.1. Forecast Comparisons First, we loo...

AI summary NSPI's 2023 load forecast projects an 825 GWh (7.3% growth) increase in total load from 2023 to 2033, driven by electrification, cooling demand, and EV adoption, offset by rooftop solar and DSM. The forecast reflects changing conditions including climate-driven fossil fuel reduction and increased electrification.

then large customers. This will be discussed in more detail in the following sections of this report. Figure 2. Firm peak demand Source: Synapse from NSPI Figure 2 and responses to Synapse IR-1 Table 2. Firm peak demand components Res Mode...

AI summary The document analyzes firm peak demand components, including residential, commercial, industrial, and DSM contributions, with a focus on load forecasting. It highlights the impact of electric vehicles and demand-side management on demand projections through 2033, noting significant residential load growth driven by electrification.

is growing the most, driven by electric vehicles, new customers, and building electrification. The commercial forecast increases at a more modest level, and the industrial load shows a small increase. Table 3. Sector energy requirements (G...

AI summary The document outlines rising energy demand across residential, commercial, and industrial sectors by 2033, driven by electrification and growth. It highlights NSPI's forecast of a 4.8% load reduction from DSM programs by 2033, while emphasizing the need to address significant increases in energy and peak requirements.

42.6 34.1 31.5 22.4 41% 2032 73.2 46.8 8.3 42.1 33.2 31.1 21.8 41% 2033 71.3 43.4 7.7 41.0 30.8 30.3 20.3 41% Source: Synapse from NSPI load forecast filings—using DSM savings from Figure 38. 2 See Section 4.6 of the forecast Report. Synap...

AI summary Synapse Energy Economics, Inc. provides evidence on NSPI’s 2023 load forecast, referencing prior Board Order recommendations and emphasizing the importance of heat pump and water heater end uses in residential energy forecasting. The text highlights ongoing forecast improvement efforts and sectorial load components.

t 6 2. ENERGY FORECAST In Table 3, we saw the sectorial components of the Nova Scotia load. Here, we will review each of them in sequence, going in the same order as in the forecast report. 2.1. Major Inputs and Regression Models In additi...

AI summary The document outlines the energy forecast methodology, emphasizing sectorial load components and reliance on the Conference Board of Canada's 20-year economic forecast. It notes uncertainties in long-term projections and the use of sensitivity analyses to assess risks.

es are generally consistent with the various bank forecasts available. However, there is greater uncertainty in any longer-term forecast, and sensitivity analyses are useful for looking at this issue. The residential model uses household i...

AI summary The analysis evaluates NSPI's residential load model, highlighting reliance on income and housing data, inconsistencies in reporting load increases, and the need to reevaluate the COVID-19 variable. The model's statistical validity is acknowledged but suggests alternative variables could improve accuracy.

effect going forward is nil, but the actual effect may decline in the future. The COVID-19 variable should be reevaluated going forward. Perhaps it should even be removed altogether from future years. The handling of DSM effects is discuss...

AI summary NSPI's load forecasting incorporates economic indicators, CBoC forecasts, and climate change adjustments (HDD/CDD trends). The approach acknowledges climate reality but notes the need for regular data updates. Economic forecast uncertainty and long-term weather trends are highlighted as key considerations.

incorporates these trends into the forecasts. This change represents an acknowledgement of climate reality and is a definite step forward. This data should be analyzed and updated on a regular basis. NSPI previously investigated the use of...

AI summary The document discusses incorporating climate trends into load forecasts, noting a 11% residential load increase (2023-2033) with DSM programs, driven by new customers and EV adoption. A regression-based SAE model includes factors like heating, cooling, and DSM savings. Additional weather data had minimal impact on forecasts.

as a result of COVID, and several binary terms to account for variances in consumption associated with specific months (i.e., time-fixed effects). 8 8 Load Forecast Report, Appendix B, pp. 1-8. Synapse Energy Economics, Inc. Evidence Regar...

AI summary The text details Synapse Energy Economics' analysis of Nova Scotia Power’s 2023 load forecast, breaking down variables like XHeat, XCool, and XOther. It explains factors influencing residential energy use, including heating/cooling degree days, appliance efficiency, and income, with projected changes of +7.4%, +60.1%, and -1.3% respectively.

ry change drivers for XOther are water heating (increased electric heater saturation), reductions in lighting use, and decreased television use. The net effect is to decrease XOther by 1.3 percent. 10 There are many factors driving the SAE...

AI summary The text discusses factors influencing load forecasts, including decreased XOther due to reduced lighting and TV use, and increased heating/cooling loads. NSPI uses the SAE model to calculate residential average consumption, adjusting for new customers, EVs, PV generation, and DSM impacts. The SAE model attributes 44% of residential use to heating, 3% to cooling, and 53% to other uses.

-9.2% -0.3% -5.6% 11.0% load Note: Res Sales = Existing Customer Load + New Customer Load + EV Load + Solar Load + RTR + DSM. Source: NSPI load forecast report Appendix B. In addition to the outsized contribution of EVs and new customer lo...

AI summary Residential load growth is driven by EVs, new customers, and electrification (heat pumps, electric water heating). Existing customer consumption is projected to rise 4.7% due to electrification and higher household compensation. Heat pump adoption is forecasted to reach 100% by 2050, with 2022 installations prompting revised projections.

rojected to decline by 0.9 percent between 2022–2032 in the 2022 load forecast analysis, the same term is projected to increase by 7.4 percent over the period 2023–2033 in the latest load forecast. 19 Meanwhile, the projected intensity for...

AI summary Load forecasts show conflicting projections: a 0.9% decline (2022–2032) versus a 7.4% increase (2023–2033). Heat pump intensity in XCool decreased in 2023 forecasts but was offset by higher regression coefficients. NSPI's models show significant discrepancies with E3's estimates, requiring further analysis using AMI data.

ates that the data collected from these pilots will be used to inform future load forecasts. 22 Load Forecast Report, p. 35. 23 Load Forecast Report, p. 37. 24 Load Forecast Report, pp. 78-79. Synapse Energy Economics, Inc. Evidence Regard...

AI summary Synapse recommends NSPI include heat pump data and consider EV load in future forecasts. NSPI excludes heat pumps due to limited adoption data, though uptake has doubled in two years. EV load is projected to grow significantly by 2033.

as penetration increases. These load management strategies should be reflected in the next load forecast with greater detail, with all assumptions supported empirically to the maximum extent possible. Solar generation As noted previously,...

AI summary NSPI's load forecast highlights rapid solar generation growth, projecting over 20-fold increases in installations. While solar may reduce summer peak loads by up to 15%, assumptions require empirical validation. Synapse recommends refining forecast models to better capture solar's impact on load profiles.

omer load has increased substantially in this year’s load forecast relative to the previous year’s forecast. New customers are expected to add 431 GWh (8.0 percent) to the residential load by 2033. 40 Synapse has previously noted concerns...

AI summary The load forecast indicates a significant increase in residential load by 2033, with new housing as a proxy for customer growth. Synapse Energy Economics, Inc. questions the validity of this proxy, citing potential issues like displacement of existing housing. NSPI's responses are deemed insufficient, and a recommendation is made to validate the proxy and monitor trends.

shortcomings of this proxy variable. Given that NSPI has increased its forecast for new customer growth, it should carefully monitor trends and make any needed modifications in the next load forecast. Other In its Evidence from 2022, Synap...

AI summary The text critiques NSPI's load forecasting methodology, particularly its use of a proxy variable for COVID-19 work-from-home impacts. NSPI has not revised its modeling approach despite a 2022 Board directive, and its use of a non-integer value for a binary variable is criticized. Residential energy usage projections increased from 2022 to 2023 forecasts, contrary to prior expectations. The Board previously recommended incorporating household demographics into load modeling.

f demand on load by time of day. 46 NSPI uses household size in its residential model calculations and has stated that it will investigate household size and demographics for a future forecast. 47, 48 Recommendations and Considerations We...

AI summary The document discusses NSPI's load forecasting methodologies for residential and commercial sectors, emphasizing the need to reassess modeling approaches for pandemic impacts and household demographics. The commercial sector is projected to grow by 6.8% over the forecast period, contrasting with a 0.58% decline in the 2022 forecast. The Board references a 2022 decision urging demographic impact analysis.

with the large general service customers? Are they implementing DSM measures to reduce load? Adding solar generation? Entering into RTR contracts? Might all this reduce their loads to some degree? Synapse Energy Economics, Inc. Evidence Re...

AI summary The industrial sector accounts for 23% of customer load, with forecasts showing a 2.7% growth increase. The Large subsector (70% of industrial load) shows minimal growth due to customer surveys and potential DSM measures. Electrification is projected to add 40 GWh by 2033, though this remains a small fraction of total industrial load.

the residential load. The DSM adjustments to the SAE model results are about half the size of the nominal savings as some savings that are driven by historical data are embedded in the model results. The residential statistical model inclu...

AI summary The document discusses adjustments to the SAE model for residential and commercial DSM savings, noting decreasing coefficients over time. Residential DSM adjustments in 2033 are -269 GWh, while commercial/industrial impacts are -205 GWh and -57 GWh. Uncertainty in forecasts is acknowledged, with recommendations to adjust factors if DSM savings increase.

lausible, although there are many uncertainties, and some aspects need refinement. There should be more discussion of the underlying factors causing peak growth and what can be done to mitigate it. 4. PEAK DEMAND RESEARCH The exploration o...

AI summary The document discusses the need for more refined analysis of peak demand growth and the importance of using interval data and AMI/LRS data to improve forecast accuracy. Sensitivity analyses are presented, highlighting temperature and economic factors as key uncertainties affecting load forecasts.

ive range than energy. However, the temperature at peak and the monthly HDD series are probably highly correlated. The economic and wind at peak components have very minimal impacts. • Figure D8 shows additional potential impacts. Especial...

AI summary The text discusses sensitivity analyses related to energy load forecasting, highlighting the impact of temperature, HDD, and policy options such as TOU/CPP. It also outlines NSPI's response to Synapse's recommendations and mentions the need for further clarifications and recommendations.

We ask that NSPI review its peak forecasting methodology in light of these discrepancies and evaluate what factors (e.g., DSM, heat pumps, weather) account for the gap (p.20). • We ask that NSPI provide a 2032 peak sensitivity analysis sim...

AI summary The text requests that NSPI review its peak forecasting methodology, evaluate factors like DSM and heat pumps, and provide a 2032 peak sensitivity analysis. It also recommends future analyses based on proactive actions to mitigate peak increases. A response from NSPI is referenced in Appendix B.

Matter No. M11108 In the Matter of Nova Scotia Power’s 2023 Load Forecast Report EVIDENCE OF JOHN D. WILSON ON BEHALF OF THE CONSUMER ADVOCATE Grid Strategies, LLC JULY 18, 2023 TABLE OF CONTENTS I. Identification & Qualifications ...........

AI summary The document outlines the 2023 Load Forecast Report by Nova Scotia Power, critiqued by John D. Wilson on behalf of the Consumer Advocate. Key issues include forecasted peak demand for EV charging, time-varying pricing impacts, and documentation quality. Implications for distribution planning are highlighted.

sses and jurisdictions, 19 design of retail rates, and performance-based ratemaking for electric utilities. 20 My professional qualifications are further summarized in Attachment 1. 21 Q: Have you testified previously in utility proceeding...

AI summary John D. Wilson, testifying on behalf of the Nova Scotia Consumer Advocate, discusses improvements to NS Power’s load forecast methods, areas needing enhancement, and implications of electrification for distribution planning.

ich load forecast 8 methods and documentation could be improved. Finally, I will discuss general implications 9 of the electrification load forecast for distribution system planning. 10 Q: Please summarize your recommendations. 11 A: I rec...

AI summary The witness recommends NS Power adopt a 0.6 kW/vehicle EV charging rate, remove time-varying pricing peak reduction benefits, update load forecast documentation, verify transformer sizing with updated EV data, and improve distribution planning for solar systems <100kW and heating electrification, particularly in small general sector areas.

small general sector. Evidence of John D. Wilson  Matter No. M11108  July 18, 2023 Page 3 1 III. Improvements Demonstrated in 2023 Load Forecast Report 2 Q: Please summarize the improvements demonstrated in the 2023 Load Forecast 3 Repor...

AI summary NS Power improved its 2023 Load Forecast Report by updating peak design temperature parameters and creating a hybrid electrification scenario. The report also evaluated using multiple weather stations but concluded that Shearwater RCS alone suffices, referencing prior evidence from 2021.

e weather stations has minimal impact 12 on the load forecast model, and that it is reasonable to continue using Shearwater RCS as 13 the only weather station informing the model.2 14 Q: Please summarize the update to the peak design tempe...

AI summary NS Power updated its peak design temperature model to use a 12-hour lagged average temperature and wind speed, concluding this improves accuracy. The expert, James F. Wilson, supported this approach based on prior evidence, while noting that relying on Shearwater RCS as the sole weather station has minimal impact on load forecasts.

meet the standard for use in the load forecast report. 15 Consensus Agreement (May 12, 2022), Matter No. M09777, Attachment A, p. 3. 16 NS Power, CA RIR-2(c), Matter No. M09777. Evidence of John D. Wilson  Matter No. M11108  July 18, 202...

AI summary The expert testifies that NS Power should exclude peak reduction benefits from time-varying pricing programs in its load forecast unless a new forecast is developed for programs effective on all peak days. The load forecast report is criticized for lacking documentation on key formulas and assumptions, with a recommendation to improve transparency in future reports.

Nova Scotia Power’s 2020 annual capital expenditure plan on behalf of the Nova Scotia Consumer Advocate. Potential to decommission hydroelectric systems, review of annually recurring capital projects, use of project contingencies, and cost...

AI summary Nova Scotia Power’s 2020 capital expenditure plan, decommissioning hydroelectric systems, and project reviews are discussed in three regulatory matters. Paul Chernick testified on behalf of the Nova Scotia Consumer Advocate in UARB matters M09579 and M09707, addressing alternatives, cost calculations, and forecast accuracy. In California PUC Docket A.19-10-012, he advocated for EV charging programs aligned with state goals, emphasizing budget controls and outreach.

3 NON-CONFIDENTIAL 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential

AI summary This document serves as a rebuttal to the 2023 Load Forecast Report, though the specific arguments or data presented in the rebuttal are not detailed in the provided text.

Page 2 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential 32 3.2.4 Recommendation 4: ............................................................................................................ 21 33 3.2.5 Recommendation...

AI summary The document outlines the rebuttal evidence for the 2023 Load Forecast Report, addressing recommendations and including a regulatory letter from NSUARB regarding Nova Scotia Power Inc.'s submission. Filed on September 14, 2023, it discusses energy usage patterns and forecasting methodologies.

Page 3 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential

AI summary The document is titled '2023 Load Forecast Report – Rebuttal Evidence' and is marked as non-confidential. It appears to be part of a regulatory proceeding, though no substantive content or arguments are present in the provided text.

1 1.0 INTRODUCTION 2 3 In accordance with the Nova Scotia Wholesale Electricity and Renewable to Retail Market Rules 1, 4 the Nova Scotia Power System Operator (NSPSO) is required each year to provide the Nova Scotia 5 Utility and Review B...

AI summary The Nova Scotia Power System Operator (NSPSO) submitted its 2023 Load Forecast Report to the Nova Scotia Utility and Review Board (NSUARB), triggering a paper hearing. Interventions were received from stakeholders including the Consumer Advocate and EfficiencyOne. Synapse Energy Economics noted improvements in the report's reasonableness and responsiveness from NS Power, while acknowledging amendments based on prior regulatory directives.

ponding directives. 25 26 Synapse noted improvements in the 2023 Load Forecast, NS Power’s responsiveness, and the 27 forecast’s reasonableness among its general introductory comments as follows: 1 Nova Scotia Wholesale and Renewable to Re...

AI summary The 2023 Load Forecast Report shows increased predictions due to electrification efforts, with implications for energy and peak demand. Synapse notes the forecast's reasonableness but highlights the need to address rising energy needs driven by population and economic growth.

several recommendations and NSPI has 18 moved to address most of them. For the record, we have attached our 19 recommendations and NSPI’s replies in attachments to this report. 20 21 Most of the issues are addressed in NSPI’s current Repor...

AI summary NSPI has addressed most recommendations, with some deferred or ongoing. The report is seen as an improvement, and NS Power made specific improvements to the load forecast, such as updating temperature and wind speed data and creating a hybrid scenario. The CA supports these efforts.

Page 5 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential 1 electrification of building heat to reflect peak mitigation through the use of non- 2 electric heat sources. 3 4 I would also like to acknowledge that NS Power e...

AI summary NS Power's 2023 Load Forecast Report incorporates improvements from prior matters, including load-weighted weather data and electrification scenarios. The report uses Shearwater RCS as the primary weather station, with the SBA noting approval of changes from the 2022 matter (Ml0569) and new data from the Smart Grid Nova Scotia report.

this Rebuttal 21 Evidence. 22 7 M11108, Exhibit N-9, Evidence of John Wilson, July 18, 2023, page 4, lines 4 - 13. 8 M11108, SBA Submission, July 18, 2023, page 1. DATE FILED: September 14, 2023 Page 6 of 25 2023 Load Forecast Report – Reb...

AI summary This rebuttal includes evidence from John Wilson's testimony (M11108, Exhibit N-9) and the SBA submission, referencing the 2023 Load Forecast Report. The document is part of a regulatory proceeding dated September 14, 2023, focusing on load forecasting and related evidence submissions.

etting peak demand increases in future years. The current TVP pilot will be concluding in 18 2024, and lessons learned from that pilot will be used to forecast future impacts from rate design. 19 DATE FILED: September 14, 2023 Page 8 of 25...

AI summary The 2023 Load Forecast Report discusses the conclusion of the TVP pilot in 2024 and its implications for future rate design. Lessons from the pilot will inform forecasting methods related to rate design impacts.

Page 8 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential

AI summary This document is page 8 of a 25-page non-confidential submission titled '2023 Load Forecast Report – Rebuttal Evidence,' indicating it is part of a regulatory proceeding involving evidence submitted in response to a load forecast report.

he sensitivities that NSPI included testing 34 this issue, and NSPI should also seek to validate any other assumptions about 35 customer usage of secondary heating equipment. 9 M11108, Exhibit N-8, Synapse Evidence, July 18, 2023, pages 28...

AI summary The text highlights the need for NSPI to validate assumptions about customer usage of secondary heating equipment in its load forecasting, citing Synapse Evidence (M11108, Exhibit N-8) from July 2023. This is part of the 2023 Load Forecast Report rebuttal evidence filed on September 14, 2023.

Page 9 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential

AI summary The document is a rebuttal submission related to the 2023 Load Forecast Report, likely challenging or providing counter-evidence to the forecast's assumptions or conclusions. No specific claims or entities are detailed in the provided text.

1 NS Power Response: 2 3 NS Power agrees with this recommendation. Heat pump data related to energy and peak effects is 4 aligned with the most recent studies available for Nova Scotia (Itron and E3). As more data is 5 gathered on the impa...

AI summary NS Power agrees to use heat pump data for load forecasting but notes data limitations. They mention developing class-level peak forecasts and addressing EV adoption with rate designs.

programmatic options to reduce on-peak EV charging as penetration increases. 31 These load management strategies should be reflected in the next load forecast with 32 greater detail, with all assumptions supported empirically to the maximu...

AI summary The NSUARB recommends refining load forecasts to include detailed, empirically validated strategies for managing EV charging and solar generation's peak reduction benefits. NS Power agrees to improve assumptions in the 2024 forecast, noting seasonal limitations in winter peak reductions and the need to validate new home construction as a customer growth proxy.

Given that NSPI has increased its forecast for new customer growth, it should 25 carefully monitor trends and make any needed modifications in the next load 26 forecast. 27 28 NS Power Response: 29 30 The new customer load forecast project...

AI summary NSPI has increased its forecast for new customer growth, prompting a recommendation to monitor trends and adjust load forecasts. NS Power responds that its annual load forecasts use data from the Conference Board of Canada, ensuring updates based on recent customer growth data.

vice customers? Are they 28 implementing DSM measures to reduce load? Adding solar generation? Entering 29 into RTR contracts? Might all this reduce their loads to some degree? 30 DATE FILED: September 14, 2023 Page 13 of 25 2023 Load Fore...

AI summary NS Power addresses questions about customer load reduction efforts, including DSM measures, solar adoption, and RTR contracts, stating impacts on large general class customers are minimal. They reference IRP scenarios for electrification and savings impacts in the 2023 Load Forecast Report.

d be adjusted upward to reflect greater levels of incremental 24 savings. 25 26 NS Power Response: 27 28 This recommendation is discussed in Section 4.6 of the Load Forecast Report. 29 DATE FILED: September 14, 2023 Page 15 of 25 2023 Load...

AI summary The NS Power responds to recommendations regarding load forecasting, clarifying that system peak calculations account for interruptible customers and demand response (DR) impacts. It notes that time-varying pricing and AMI-enabled strategies are already factored into peak forecasts, with updates pending Smart Grid and pilot data. Heat pump performance during peak conditions is under investigation.

EV peak 9 impact. 10 11 3.2.2 Recommendation 2: 12 13 NS Power should remove the peak reduction benefits associated with time-varying 14 pricing from its load forecast. 15 16 NS Power Response: 17 18 NS Power agrees that the forecast TVP p...

AI summary NS Power acknowledges overstated TVP peak savings in 2023-2024 but disputes removing them from load forecasts, citing minimal impact. The CA argues pilot programs lack coverage of all peak periods, while NS Power suggests flexibility in CPP events could improve alignment. NS Power will revise 2024 forecasts based on updated uptake estimates.

Page 19 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential 1 with times where those savings are most required (i.e. times of high ANL where non-dispatchable 2 generation is low). To address this issue, in the TVP Pilot Pr...

AI summary NS Power argues that demand response from TVP rates should be evaluated using Equivalent Load Carrying Capability (ELCC) adjustments, as load forecasts only estimate system-level peak demand. They recommend modifying the TVP tariff to include weekend events and a sliding 4-hour window to better align with peak demand periods.

rates. 19 20 3.2.3 Recommendation 3: 21 22 NS Power should update its load forecast documentation to include the additional 23 details supplied in response to CA IR-10. 24 25 NS Power Response: 26 27 NS Power agrees with this recommendatio...

AI summary The Board recommends NS Power update its load forecast documentation with details from CA IR-10, specifically HeatUse, CoolUse, and OtherUse variables. NS Power agrees and will include these in the 2024 Load Forecast Report. A cross-reference to M09777, the Time Varying Pricing Pilot Program report, is cited.

also recommends a review of historical load survey sales realization for 27 survey-based class sales forecasts… 28 12 M11108, SBA Submission, July 18, 2023, page 3. DATE FILED: September 14, 2023 Page 22 of 25 2023 Load Forecast Report – R...

AI summary The text recommends reviewing historical load survey data for accuracy in sales forecasts and references a Small Business Advocate submission (M11108) regarding the 2023 Load Forecast Report rebuttal evidence.

Page 22 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential 1 NS Power Response: 2 3 NS Power agrees with this recommendation and will provide a comparison in the 2024 Load 4 Forecast Report. 5 6 3.3.3 Recommendation 3: 7...

AI summary NS Power agrees with the SBA's recommendation to address RTR load impacts in future reports and acknowledges the NSUARB's direction to adjust for an unexpected load increase from a customer switching back to NS Power services.

an increase in load for 2023. 30 The Board directs NS Power to address this increase to load in its Rebuttal Evidence 31 to be filed with the Board by September 14, 2023. 13 13 NSUARB Letter re: M11108 – Nova Scotia Power Inc. - 2023 Load...

AI summary The NSUARB directed NS Power to address a 2023 load increase in its rebuttal evidence by September 14, 2023. NS Power responded that updating forecasts for isolated assumptions lacks meaningful clarity and plans to address RTR changes in the 2024 Load Forecast report.

Page 24 of 25 2023 Load Forecast Report – Rebuttal Evidence Non-Confidential 1 4.0 CONCLUSION 2 3 Submission of the Load Forecast Report is an annual requirement under the provisions of the Nova 4 Scotia Wholesale and Renewable to Retail M...

AI summary NS Power submits its 2023 Load Forecast Report as required under Nova Scotia's market rules, highlighting methodological improvements and stakeholder input since 2007. The report incorporates feedback and will be adjusted based on the NSUARB's determination. NS Power requests acceptance of the report.

November 7, 2023 [email protected] Mark Peachey Manager, Capital Filings Nova Scotia Power 1223 Lower Water St Halifax, NS B3J 3S8 Dear Mr. Peachey: M11108 – Nova Scotia Power Inc. – 2023 Load Forecast Report (P-194) Nova Scotia Powe...

AI summary NS Power filed its 2023 Load Forecast Report, detailing energy and peak demand projections from 2023 to 2033. The Board established a paper hearing timeline, approved confidentiality, and received interventions from groups like the Industrial Group and EfficiencyOne. Synapse Energy Economics reviewed the filing, and the Board directed NS Power to address an unplanned service continuation in its rebuttal evidence.

rowth of 2.3% from 2024 to 2033. This growth is attributed to customer growth and increased electrification of heating and vehicles, which will be mitigated by DSM and Demand Response (DR) activities. These forecasts project the 2033 NSR w...

AI summary NS Power forecasts a 2.3% energy demand growth from 2024 to 2033, driven by customer growth and electrification of heating/vehicles, with DSM and Demand Response (DR) expected to mitigate increases. The 2033 NSR is projected to rise 8.8% (979 GWh) above 2022 levels, with peak demand increasing 27.2% (603 MW). Table 1 compares forecasted and actual energy and peak demand data from 2018-2023.

ased on Table A1 in each annual report starting with prior year actual 2 Based on Table A2 in 2022 and 2021 Load Forecast reports and in Table A3 in each annual report from 2015 to 2020 and 2023 Document: 308584 -3- The load forecast is a...

AI summary The document discusses NS Power's load forecasting practices, emphasizing their critical role in planning and budgeting. The Board has raised concerns about forecast accuracy and consistency, prompting NS Power to engage stakeholders and implement improvements like enhanced weather data modeling and sensitivity analyses following the 2022 Board Decision (M10569).

NS Power update the distribution system planning methods and practices to account for the impacts of small scale solar and electrification of heating loads, particularly for the small general sector. The SBA raised concerns about forecasti...

AI summary NS Power seeks to update distribution planning methods for small-scale solar and heating electrification. The SBA raised concerns about forecasting accuracy, RTR load impacts, and EV adoption rates. Synapse recommended improvements to load forecasting, including empirical validation and scenario modeling.

he peak forecasting method; and • performing future sensitivity analysis based on actions to reduce projected peak increases using newly available technology. NS POWER REBUTTAL EVIDENCE In its rebuttal evidence, NS Power addressed the conc...

AI summary NS Power addresses intervenors' concerns about peak forecasting methods, confirming its approach to EV charging impacts and transformer sizing. The company uses AMI data and customer demographics to model distribution systems, while responding to Synapse's recommendations on residential peak forecasts and heat pump integration, noting data limitations for bottom-up forecasting.

recast will be developed, but a bottom-up end-use specific forecast is not possible because NS Power lacks end-use load data. Document: 308584 -5-

AI summary The development of a recast forecast is underway, but a detailed bottom-up end-use specific forecast is not feasible due to NS Power's absence of end-use load data, which is critical for such analysis.

NS Power identified that real time pricing is available to industrial customers but only used by one. Other industrial rates are being developed for hydrogen production facilities, however, NS Power regards testing of rate design as outsid...

AI summary NS Power acknowledges real-time pricing for industrial customers but notes limited adoption. They agree to incorporate data from the Integrated Resource Plan and TVP Pilot into forecasts but disagree with removing TVP savings. NS Power will refine EV load assumptions and include solar generation impacts in the 2024 forecast, though they contest applying a fixed EV load increase due to variable charging data.

priate because the current sample group charging data shows that EV load demand is variable. Therefore, a better understanding of EV’s effect on peak is needed before results from the sample can be confidently applied to the forecast. The...

AI summary The Board highlights concerns about EV load variability and the need for a low-adoption electrification scenario to assess system needs and DSM benefits. NS Power's load forecast is criticized for not addressing isolated assumption changes, while the Board emphasizes stakeholder engagement for the 2024 Load Forecast Report. Intervenors' recommendations on load forecasting methodology are emphasized.

are important to discuss in the initial stakeholder meeting. It is helpful if NS Power shares how potential impacts were evaluated and what was incorporated into the forecast early in the process. Document: 308584 -6- The Board notes that...

AI summary The Board acknowledges NS Power's agreement with Intervenor's recommendations and directs the implementation of specific Load Forecast improvements, including IRP, AMI, TVP Pilot outcomes, and carbon emission model reviews. It encourages evaluating model elasticity and input variables for robustness.

wer System Planning Team will conduct more resource modelling “…to assess the potential load impacts and the balance of energy and capacity requirements to meet the net demand of hydrogen facilities.” The Board agrees the issue of commerci...

AI summary The Board emphasizes addressing commercial hydrogen production's impact on NS Power's grid, noting self-generation risks and the need for updated load forecasts. It highlights a 234 GWh variance between 2022 forecasts and actual Net System Requirements, attributing it to pandemic and heat pump factors, and urges NS Power to revisit this category.

M11108 NOVA SCOTIA UTILITY AND REVIEW BOARD IN THE MATTER OF an application by NOVA SCOTIA POWER INCORPORATED for approval of 2023 Load Forecast Report LIST OF PARTICIPANTS NOVA SCOTIA POWER INC. (NS Power) Mark Peachey 1223 Lower Water St...

AI summary The Nova Scotia Utility and Review Board is considering an application by Nova Scotia Power Inc. for approval of its 2023 Load Forecast Report. Key participants include NS Power, Board Counsel William J. Mahody, and Synapse Energy Economics, Inc. as a consultant. The proceeding involves regulatory approval of forecasting methodologies.

Document #:303618 Date Filed: May 25, 2023 Synapse (NSPI) Page 1 of 18 1 Questions regarding the NSPI “2023 Load Forecast Report” of April 28, 2022 2 Request IR-1: 3 Report Tables 4 a. Please provide in electronic spreadsheet format all ta...

AI summary The document outlines three requests (IR-1, IR-2, IR-3) for data related to NSPI's 2023 Load Forecast Report, including spreadsheet formats for tables, historical sales data, and clarification on billed vs. accrued sales methodologies. The focus is on transparency and accuracy in load forecasting and historical energy usage.

ndustrial “to improve the fit.” Please quantify the nature of the improved 28 fit. 29 c. Please identify the effect if the industrial forecast used the same period as the residential 30 and commercial forecasts (that is, January 2013 to De...

AI summary The text outlines requests for detailed weather data and methodology related to heating degree days (HDD) and cooling degree days (CDD) used in a forecast. It asks for temperature data, calculation methods, and explanations for data period changes, focusing on accuracy and consistency in energy demand modeling.

h. Regarding the weighing of weather station data and the comparisons between the two 15 approaches which is presented as the Mean Absolute Percentage Error (MAPE)? Were 16 other comparison approaches investigated? Why was the MAPE method...

AI summary The document contains requests for detailed economic data and methodology explanations from Nova Scotia Power Incorporated (NSPI), including data sources, model construction, and assumptions related to population forecasts, housing completions, and sector-specific GDP/employment variables in their economic models.

Document #:303618 Date Filed: May 25, 2023 Synapse (NSPI) Page 6 of 18 1 End-Use Intensities (Section 4.4, pp. 47-50) 2 a. Please provide the underlying calculations for the residential changes in baseboard 3 heating, heat pump heating, co...

AI summary The document contains regulatory requests for detailed calculations and explanations regarding end-use intensities, commercial/industrial growth forecasts, and load forecasting methodologies. It seeks clarification on residential and commercial energy use changes, heat pump impacts, and compliance with prior Board directives on economic inputs and model elasticity.

Document #:303618 Date Filed: May 25, 2023 Synapse (NSPI) Page 8 of 18 1 d. Please describe any changes in the data and statistical analysis used to develop the 2 coefficient for the DSM variable for the Commercial and Industrial forecast...

AI summary The document contains regulatory requests (IR-18) directed at NSPI, seeking clarifications on statistical methods for DSM coefficients, residential solar adjustments, and the impact of COVID-19 on energy forecasts. Requests focus on data sources, calculation methodologies, and assumptions related to long-term work-from-home trends.

e impacts of COVID on the 2021 and 2022 loads. 12 b. Please explain and quantify the ongoing effects of COVID in the commercial forecast. 13 c. Please explain why the current forecast (Figure 44) remains shows an increase through 14 2033 w...

AI summary The text outlines regulatory requests (IR-21 to IR-23) seeking explanations for forecast discrepancies in electricity demand, focusing on factors like COVID impacts, EV loads, space heating, DSM programs, and solar generation. Requests emphasize quantifying model changes, evaluating efficiency trends, and clarifying forecast assumptions for 2021–2033.

ease explain and quantify the specific reasons for the differences from the previous 5 forecast. 6 b. Please quantify how much of the increase is associated with expansion of institutional 7 facilities and how much from other factors. 8 9...

AI summary The document outlines requests for detailed explanations and quantifications regarding forecast discrepancies, growth factors, survey representation, municipal load changes, and system losses. It seeks data on industrial and municipal energy usage trends, reliability of projections, and historical system loss patterns.

s and explain. 26 27 Request IR-33: 28 Sensitivity Analysis (Section 11, 2020 IRP Comparison, pp 97-98) 29 a. Please provide information about how the evergreen IRP analysis might affect future 30 loads. 31 b. For comparison with the previ...

AI summary The text includes two requests for information: IR-33 asks about the impact of the evergreen IRP analysis on future loads and seeks 2032 data from Figure 68 for comparison with prior forecasts. IR-34 is partially listed but incomplete.

ow the evergreen IRP analysis might affect future 30 loads. 31 b. For comparison with the previous forecast please provide values for 2032 in Figure 68. 32 33 34 Request IR-34:

AI summary The text references an 'evergreen IRP analysis' and requests comparative 2032 data from Figure 68. It focuses on forecasting methodology and the impact of integrated resource planning on future load projections, though no specific claims or entities are explicitly mentioned.

lain why Manufacturing Employment was chosen as the long-term driver for 33 this forecast, whether any other variables were considered, and why they were rejected. 34 35 Request IR-40:

AI summary The text references a request (IR-40) seeking justification for selecting Manufacturing Employment as the long-term driver in a forecast, including consideration and rejection of alternative variables. The focus is on forecasting methodology and variable selection rationale.

Document #:303618 Date Filed: May 25, 2023 Synapse (NSPI) Page 15 of 18 1 Appendix B: Combined Model for Commercial and Industrial DSM Coefficient (pp 27-28) 2 a. Please provide in electronic spreadsheet format the data and the statistical...

AI summary The document contains regulatory requests (IR-41 to IR-43) directed at Synapse (NSPI), seeking detailed data, model parameters, and explanations for DSM coefficient models, peak forecasts, and sensitivity analyses. Requests focus on transparency of methodologies, data sources, and assumptions used in NSPI's submissions.

tronic form. 28 b. Please provide the full set of Oracle input data in electronic format so that it can be 29 replicated. 30 c. Please provide the historical data and forecasts used to develop the probability 31 distributions for each vari...

AI summary The document includes requests for data and reports related to electrification support and forecast improvements. Nova Scotia Power (NSP) committed to addressing intervenor concerns by incorporating multi-station weather analysis, peak sensitivity studies, and electrification impact assessments into future forecasts.

Document #:303618 Date Filed: May 25, 2023 Synapse (NSPI) Page 17 of 18 1 Request IR-46: 2 General Forecast Report Improvements – The Board Decision of October 31, 2022 Includes the 3 following findings (pp 5-6): 4 The Board notes that NS...

AI summary The Board directed NS Power to improve its Load Forecast by detailing electrification and technology impacts, evaluating economic model inputs (including DSM, Solar PV, EVs, and weather), and enhancing stakeholder engagement. Intervenors' recommendations on weather-normalized values and stakeholder collaboration were emphasized.

prudent that NS Power 23 share how potential impacts were evaluated and what was incorporated into the forecast 24 early in the process. 25 26 In addition to the above directives, the Board encourages NS Power to include 27 information on...

AI summary The Board requests NS Power to evaluate and incorporate specific factors into load forecasts, including elasticity models, housing completions, income metrics, and household demographics. The directives aim to improve forecast accuracy and transparency.

M11108 NOVA SCOTIA UTILITY AND REVIEW BOARD IN THE MATTER OF: THE PUBLIC UTILITIES ACT - and - IN THE MATTER OF: NOVA SCOTIA POWER INCORPORATED (NS Power) 10 – Year Energy and Demand Forecast (2023 Load Forecast Report) INFORMATION REQUEST...

AI summary The Nova Scotia Utility and Review Board has issued an information request to Nova Scotia Power Inc. regarding their 10-year energy and demand forecast (2023 Load Forecast Report), seeking responses by June 20, 2023, with contact details provided for follow-up.

Document: 303590 (P-194) Date Filed: 05/30/23 UARB Page 1 1 Request IR-1: 2 Pages 16 and 17 of the Application discuss customers taking service under the Renewable to 3 Retail (RTR) Tariff in 2023. Lines 3 and 4 on page 17 state: “…the cor...

AI summary The document contains four requests (IR-1 to IR-4) questioning Nova Scotia Power's (NSP) preparedness for RTR tariff adoption delays or higher take-up, the use of a single HDD reference temperature (18°C), housing start correlations with customer forecasts, and inflation-adjustment methodologies for financial variables. Each request seeks clarification on assumptions and data practices.

dollars to eliminate inflation effects. Have all the variables back to 2013 been adjusted? 28 b) Figure 17 Residential Economic Drivers provides the number of New Construction with a 29 growth rate of 35.5% in 2023. Is this data provided b...

AI summary The text raises questions about data adjustments for inflation, verification of new construction growth rates by the Canadian Mortgage Housing Corporation, and the impact of revised 2022 Load Forecast data on load projections. It also inquires about declining household compensation trends.

eport. Please explain why and describe how does this revised 33 data effect the load forecast? 34 c) Figure 17 provides Household Compensation which shows a decline in 2023.

AI summary The text requests an explanation of how revised data impacts load forecasting and highlights a decline in household compensation in 2023 as shown in Figure 17. The focus is on understanding the implications of updated data for forecasting accuracy and identifying trends in compensation metrics.

Document: 303590 (P-194) Date Filed: 05/30/23 UARB Page 2 1 i. The figures for 2020 onward have been revised from the 2022 Load Forecast 2 report. The 2022 report forecasted compensation in 2022 at 18,637 and 2023 at 3 18,747. This report...

AI summary The document requests explanations for revised load forecast data (2022-2023 compensation, non-manufacturing employment) and justification for 2023 GDP/employment trends. It questions data sources (e.g., Conference Board of Canada), cross-checking with other forecasts, and the impact of revised figures on load projections.

If so, please 20 provide a copy. 21 iii. Has the data been checked against other forecasts? 22 e) Figure 19: Industrial Economic Drivers provides the historical and forecast manufacturing 23 GDP and employment both declining in 2023. 24 i....

AI summary The text requests clarification on revised manufacturing GDP and employment data from the 2022 Load Forecast report, including reasons for negative 2023 growth, discrepancies between employment and GDP trends (2025–2027), and validation against other forecasts. It emphasizes the impact of these revisions on load forecasting.

Document: 303590 (P-194) Date Filed: 05/30/23 UARB Page 5 1 Request IR-12: 2 Page 59 of the Application states in lines 5 to 7 “Building efficiency is expected to improve, 3 although based on calibration done in 2013, improvements are expe...

AI summary The text includes six requests (IR-12 to IR-16) questioning the Application's use of outdated building efficiency data, discrepancies in new customer load forecasts, EV energy allocation assumptions, economic downturn considerations, and the inclusion of a COVID variable in modeling. Requests seek clarification on methodology, data sources, and justification for projections.

Document: 303590 (P-194) Date Filed: 05/30/23 UARB Page 6 1 Request IR-17: 2 Page 65 of the Application states in line 13 “the addition of EV loads adds 272 GWh by 2023….” 3 Please explain how the EV load has been verified. 4 5 Request IR-...

AI summary The UARB requests clarification on NSP's load forecasts, including EV load verification, confidence in hospital project growth, economic growth assumptions for industrial sales, metrics supporting industrial decline, and confidence in new industrial project forecasts. Questions focus on forecasting methodology, DSM measures, and data validation.

confidence in this forecast level of new project energy requirements? 30 b) Have high and lows been estimated? 31 c) Have any DSM or other energy mitigation measures been incorporated?

AI summary The text presents three questions regarding energy forecasting: confidence in projected energy requirements, estimation of high/low scenarios, and inclusion of demand-side management (DSM) or other mitigation measures in the forecast.

Document: 303590 (P-194) Date Filed: 05/30/23 UARB Page 7 1 Request IR-22: 2 Figure 53 of the Application indicates that under the Residential Class, 263 GWh are the result of 3 an unexplained variance from the 2022 forecast. This is a sig...

AI summary The document contains four regulatory requests (IR-22 to IR-25) questioning Nova Scotia Power's (NSP) forecasting methodology, including unexplained energy variances, changes in Net System Requirement (NSR) forecasts, residential component allocations, and the incorporation of hydrogen production into load forecasts. Requests seek clarification on factors driving discrepancies, forecast adjustments, and data gaps related to hydrogen's impact.

1 M11108 2 3 NOVA SCOTIA UTILITY AND REVIEW BOARD 4 5 IN THE MATTER OF: The Public Utilities Act 6 7 -and - 8 9 IN THE MATTER OF: an application by NOVA SCOTIA POWER INCORPORATED 10 for approval of its 2023 Load Forecast Report 11 12 13 14...

AI summary The Nova Scotia Consumer Advocate has issued an information request to Nova Scotia Power Inc. regarding its 2023 Load Forecast Report, seeking responses by June 20, 2023. The request is part of a proceeding under the Public Utilities Act, with details on contact persons and filing requirements.

Date Filed: May 30, 2023 CA (NS Power) Page 2 of 6 1 (a) Please provide the system-coincident unmanaged peak impact per vehicle for 2022. In 2 other words, perform the same calculation for the system peak hour or, ideally, for the 3 top te...

AI summary The document includes requests to NS Power regarding EV charging impact calculations, distribution planning considerations for EV and solar growth, and verification of TVP's exclusion from the 2023 Load Forecast. It seeks clarification on AMI technology's timeline for reducing system peak demand and examines peak event timing during winter periods.

Date Filed: May 30, 2023 CA (NS Power) Page 4 of 6 1 In his 2022 evidence, Mr. Wilson pointed out potential modeling errors in the residential 2 energy model. Mr. Wilson noted that the residential WtXHeat output did not exhibit the trend 3...

AI summary The text requests NS Power to confirm modeling accuracy in residential and small general load forecasts, address critiques from Mr. Wilson, and explain changes between 2022 and 2023 models. It also references Exhibit N-1 for a calculation example and seeks clarification on NS Power's responses to prior critiques.

1 M11108 2 3 NOVA SCOTIA UTILITY AND REVIEW BOARD 4 5 IN THE MATTER OF: The Public Utilities Act, R.S.N.S. 1989, c.380, as amended 6 7 - and - 8 9 IN THE MATTER OF: 2023 Load Forecast Report 10 11 12 13 INFORMATION REQUESTS 14 15 16 17 To:...

AI summary The Nova Scotia Utility and Review Board issues an information request to Nova Scotia Power Inc. under the Public Utilities Act, seeking responses by June 20, 2023, regarding the 2023 Load Forecast Report. The request is issued by the NS SBA with contact details provided.

Page 1 of 3 1 Request IR-1: Please reference Synapse IR-3 parts A-C, and reference Section 4.0, p 16. 2 a) Please provide a qualitative description of the changes in any class studied that would 3 merit differentiation in regression period...

AI summary The text contains four information requests (IR-1 to IR-4) seeking details on regression periods, load forecasting assumptions, EV charging data alignment, and Commercial Net Metering protocols. Requests focus on methodology, data benchmarks, model accuracy, and protocol integration in load/peak forecasts.

M11108 – SBA IRs – May 30, 2023 Page 2 of 3 1 b) Please describe how the RTR-served load/generation is forecasted. 2 c) Please provide the total amounts of load per year and per rate class to be served by RTR 3 providers in all years of th...

AI summary The document contains information requests from a regulatory proceeding, asking NS Power to detail load forecasting methods, EV sales assumptions, and BNI curtailment programs. It seeks clarification on forecasting methodologies, EV sales alignment with federal targets, and the role of E1 in BNI curtailment.

ow much of this BNI curtailment is done by programs provided by E1? Please describe 26 all the programs E1 provides that contribute to this curtailment forecast. 27 c) Please discuss in detail and provide data supporting the BNI curtailmen...

AI summary The text includes requests for information on BNI curtailment programs, historical peak predictions, and integration of extreme scenarios in NS Power's planning. Questions focus on data supporting BNI curtailment forecasts, regression model accuracy, and system planning for extreme load scenarios.

Figure 23 23 Please provide a version of Figure 23 that includes every year of the load forecast (2023-2033). 24 25 Request IR-04: 26 Reference: NS Power 2023 Load Forecast, Page 40, Lines 7-8 Date Filed: May 30, 2023 E1 (NS Power) Page 2...

AI summary EfficiencyOne (E1) requests a revised Figure 23 from Nova Scotia Power (NSP) to include the full 2023-2033 load forecast period, citing NSP's 2023 Load Forecast Report (Page 40, Lines 7-8) as the basis for the request. The matter is part of regulatory proceedings under M11108.

ower 2023 Load Forecast, Page 45, Figure 31 25 Please provide a version of Figure 31 that breaks out total new installs, load (GWh) and peak 26 (MW) by residential and non-residential projects. 27 Date Filed: May 30, 2023 E1 (NS Power) Pag...

AI summary EfficiencyOne (E1) requests detailed information from Nova Scotia Power Inc. (NS Power) regarding the 2023 Load Forecast, focusing on how paused commercial net metering projects (over 27 kW) are accounted for, the number of paused projects, and integration of the new Commercial Net Metering Program (up to 1 MW) into the forecast.

Blackburn Law July 18, 2023 VIA EMAIL Ms. Crystal Henwood Regulatory Affairs Officer/Clerk Nova Scotia Utility and Review Board 1601 Lower Water Street, 3rd Floor Halifax NS B3J 3S3 Dear Ms. Henwood: Re: M11108 - Nova Scotia Power Inc. ("N...

AI summary The Small Business Advocate (SBA) reviews Nova Scotia Power Inc.'s 2023 Load Forecast Report, noting improvements from the 2022 version but expressing concerns about the methodology's lack of scenario analysis for electrification. The SBA highlights incorporated changes like load-weighted weather data and hybrid electrification scenarios but emphasizes the need for more comprehensive energy need assessments.

tes 827 GWh of EV load by 2032, whereas the 2022 edition only included 510 GWh by 2032 4 • Peak load has nearly doubled from a 2022 value of 89 MW in 2032 to a 2023 value of 173 MW in 2032 5 • NS Power notes that the 2022 Load Forecast dem...

AI summary NS Power's 2023 load forecast shows increased commercial sales due to EV allocation and higher peak loads, but lacks assumptions about electrification incentives. The SBA warns that excluding current incentive impacts risks overestimating electrification's short-term effects, potentially misguiding IRP capital planning and rate class allocations. The SBA recommends including a lower electrification scenario in analyses.

November 7, 2023 [email protected] Mark Peachey Manager, Capital Filings Nova Scotia Power 1223 Lower Water St Halifax, NS B3J 3S8 Dear Mr. Peachey: M11108 – Nova Scotia Power Inc. – 2023 Load Forecast Report (P-194) Nova Scotia Powe...

AI summary NS Power filed its 2023 Load Forecast Report, outlining energy and peak demand projections through 2033. The Board established a hearing timeline, approved confidentiality, and received interventions from groups like the Industrial Group and EfficiencyOne. Synapse reviewed the filing, and the Board adjusted deadlines for evidence submissions. NS Power was directed to address an unplanned service continuation in its rebuttal.

rvice in its Rebuttal Evidence. NS Power filed its Rebuttal Evidence on September 14 ,2023. Document: 308584 -2- 2023 Load Forecast NS Power uses two discrete elements for its forecast. The first is the Statistically Adjusted End- Use (SAE...

AI summary NS Power's 2023 Load Forecast projects near-term growth driven by customer and EV adoption, with mid-term growth offset by DSM and DR initiatives. The forecast combines SAE models, industrial econometric data, and DSM adjustments to estimate net system energy requirements (NSR) and peak demand growth.

rowth of 2.3% from 2024 to 2033. This growth is attributed to customer growth and increased electrification of heating and vehicles, which will be mitigated by DSM and Demand Response (DR) activities. These forecasts project the 2033 NSR w...

AI summary Energy demand in Nova Scotia is projected to grow by 2.3% from 2024 to 2033 due to customer growth and electrification, offset by DSM and DR initiatives. Forecasts show 8.8% higher NSR and 27.2% greater peak demand by 2033 compared to 2022, with Table 1 comparing forecasted vs actual energy and peak demand data from 2018 to 2023.

ased on Table A1 in each annual report starting with prior year actual 2 Based on Table A2 in 2022 and 2021 Load Forecast reports and in Table A3 in each annual report from 2015 to 2020 and 2023 Document: 308584 -3- The load forecast is a...

AI summary The document emphasizes the critical role of load forecasting in NS Power’s planning and operations, noting past Board concerns about accuracy. It references the 2022 Board Decision (M10569) and stakeholder consultations addressing forecasting methods, including temperature modeling, solar adjustments, and Smart Grid project integration.

me in the residential model; • adjusting for small scale solar panel installations; • conducting additional sensitivity analyses; and, • incorporating the results from the Smart Grid project. The Load Forecast was adjusted for the RTR load...

AI summary NS Power adjusted the 2023 Load Forecast by subtracting 14 GWh starting in 2024, while maintaining peak forecasts to meet legal obligations. Variances in 2022 were attributed to weather, the pandemic, and heat pumps. Intervenors praised forecast improvements but emphasized researching electrification impacts and new technologies like Smart Grid and Time Varying Pricing.

NS Power update the distribution system planning methods and practices to account for the impacts of small scale solar and electrification of heating loads, particularly for the small general sector. The SBA raised concerns about forecasti...

AI summary NS Power seeks to update distribution planning methods for small-scale solar and heating electrification. The SBA raised concerns about forecasting accuracy, customer survey reliability, and RTR load impacts. Synapse highlighted improvements but recommended scenario modeling, empirical validation, and sensitivity analysis for future forecasts.

he peak forecasting method; and • performing future sensitivity analysis based on actions to reduce projected peak increases using newly available technology. NS POWER REBUTTAL EVIDENCE In its rebuttal evidence, NS Power addressed the conc...

AI summary NS Power's rebuttal evidence addresses intervenors' concerns regarding peak forecasting methods, EV charging load modeling, and transformer sizing. The company confirmed reliance on current EV charging data, plans to use AMI meter data for distribution modeling, and acknowledged limitations in developing detailed end-use forecasts due to data gaps.

recast will be developed, but a bottom-up end-use specific forecast is not possible because NS Power lacks end-use load data. Document: 308584 -5-

AI summary A forecast (recast) will be developed, but a bottom-up end-use specific approach is not feasible due to NS Power's lack of end-use load data, highlighting limitations in forecasting methodology.

NS Power identified that real time pricing is available to industrial customers but only used by one. Other industrial rates are being developed for hydrogen production facilities, however, NS Power regards testing of rate design as outsid...

AI summary NS Power agrees to update load forecasts with data from IRP, Smart Grid NS, and TVP Pilot, while disagreeing on removing TVP savings due to peak demand alignment. They will assess EV, battery storage, and heat pump impacts but contest applying a fixed 0.6 kW EV load increase due to variable charging data.

priate because the current sample group charging data shows that EV load demand is variable. Therefore, a better understanding of EV’s effect on peak is needed before results from the sample can be confidently applied to the forecast. The...

AI summary The Board emphasizes the need for better understanding of EV load variability and system impacts under low electrification adoption scenarios. It directs NS Power to address RTR service withdrawal concerns and improve stakeholder engagement in load forecasting. The Board also highlights the importance of evaluating DSM benefits and ensuring service reliability during generation shortfalls.

model. • Evaluate the input variables in the residential model and test, over a period of time, if alternative inputs make the residential model more robust, considering the following: o Given the continued population growth in Nova Scotia...

AI summary The text discusses evaluating residential model variables for robustness, considering factors like population growth, household demographics, economic data, and EV adoption rates. It also addresses the impact of Nova Scotia's hydrogen strategy on load forecasting, with NS Power committing to collaborate with stakeholders and update forecasts accordingly.

wer System Planning Team will conduct more resource modelling “…to assess the potential load impacts and the balance of energy and capacity requirements to meet the net demand of hydrogen facilities.” The Board agrees the issue of commerci...

AI summary The Board emphasizes addressing commercial hydrogen production, noting Everwind Fuels' wind farm plans for green ammonia. NS Power must model grid impacts of self-generated hydrogen power and address forecast variances linked to pandemic and heat pump effects. Load forecasting and system reliability are highlighted as critical areas for NS Power's analysis.