Topic:"Forecasting Methodology" in M12661

1 VI. FORECASTED ENERGY REQUIREMENTS 2 Q WHAT IS NS POWER PROPOSING IN TERMS OF PHP'S FORECAST ENERGY 3 REQUIREMENTS IN 2026 AND 2027 IN CALCULATING PHP'S NEW ABOVE-THE-

AI summary NS Power is proposing forecasted energy requirements for Port Hawkesbury Paper LP (PHP) in 2026 and 2027 as part of the regulatory proceeding, focusing on calculating PHP's new above-the-line energy needs.

26 Q. How do the forecast energy requirements included in NS Power's proposed 27 Above-the-Line Tariff compare to PHP's current forecast? 28 29 A. PHP's annual forecast energy requirements based on the most recent information 30 provided t...

AI summary PHP's forecast energy requirements for 2026 and 2027 (774,000 MWh annually) are lower than those used in the GRA, due to fluctuating order books and challenges in forecasting beyond one year in the paper market. PHP Wind's generation is forecast at 154,610 MWh (2026) and 507,400 MWh (2027).

NSPI Responses to Industrial Group Information Requests 1 Request IR-16: 13 From experience, the Company expects that PHP will, at times, deviate from the optimal 14 forecast. These deviations are accounted for in the ELIADC Tariff under C...

AI summary NSPI explains that deviations in PHP forecasts are accounted for in the ELIADC Tariff under specific cause codes. These deviations affect the forecasted benefits of the ELID Tariff and are analyzed in Synapse IR-16. Additionally, ELID Tariff consumption is separate from the Goose Harbour Lake Wind Farm's generation, which is treated like any other Independent Power Producer (IPP).

Part B – Protocol Forecasting and Operation - (1) Annually, no later than the seventh business day of November, NS Power will forecast the Monthly Demand Schedule, Weekly Demand Schedule, and monthly and weekly limits based on PHP's demand...

AI summary NS Power must annually forecast demand schedules by November 7th, updating them as needed, and provide daily seven-day schedules optimized in system planning. PHP's demand is co-optimized with NS Power's portfolio, with results published in ADC Schedules 1 & 2.

s PHP's energy forecast for the test years in Section V at page 27, line 2 to page 28, line 9. PHP's testimony discusses its updated forecast energy requirements at page 9, line 26 to page 10, line 7. c) The dollar value of the Interruptib...

AI summary The text discusses PHP's energy forecasts, the Interruptible Credit for interruptible load, Active Demand Control (ADC) cost recovery proposals, and the Dispatchable Rider. PHP and NS Power will analyze ADC's value for a proposed rate rider, while BAI addresses interruptible credit calculations. The Dispatchable Rider's role in capturing load flexibility is highlighted.

Part B – Protocol Forecasting and Operation - (1) Annually, no later than the seventh business day of November, NS Power will forecast the Monthly Demand Schedule, Weekly Demand Schedule, and monthly and weekly limits based on PHP's demand...

AI summary NS Power must annually forecast demand schedules for PHP and NSPSO, updating them as needed. Daily Seven Day Demand Schedules are optimized within NS Power's system planning, with PHP's demand co-optimized in the broader portfolio. Forecasts bound daily operations for Seven Day and Intra Day schedules.

3.2 PHP EVIDENCE PHP states in their evidence that they do not agree with certain aspects of NSP's proposed tariff, and specifically: - 1. the use of a 57,000 kW winter month system coincident demand for PHP interruptible load; - 2. the va...

AI summary PHP disputes NSP's proposed tariff, challenging the 57,000 kW winter demand assumption, the Interruptible credit value, the R/C ratio for a new Above-the-Line Tariff, and the 2026-2027 forecast energy requirements. These issues relate to rate design, revenue modeling, and forecasting accuracy.

4.3 VALUE OF INTERRUPTIBLE CREDIT AND UPDATES TO ENERGY SALES FORECASTS PHP's consultant also recommends changes to the interruptible credit to be priced at NSP's marginal cost of capacity and updating the 2026 and 2027 PHP sales forecasts...

AI summary The document discusses NSP's proposal to set the interruptible credit at marginal cost of capacity, conflicting with PHP's argument for a higher rate of $13.107/kVA. NSP cites settlement agreement terms, while InterGroup supports NSP's position. The LIIR rate is noted as lower than NSP's historical practice for interruptible credits.

rmation Requests from intervenors (related to sales and generation forecasting, rate design, rate impacts, capital planning, revenue requirement, etc.). Managed projects on development and submission of 2012/14, 2016/19, 2022/23 and 2025/2...

AI summary The text outlines activities related to managing rate applications for NTPC and QEC, providing regulatory support, economic analysis, and load forecasting. Key focus areas include rate design, capital planning, revenue requirements, and fuel stabilization fund rider applications.

- 6 A. As is shown in [Table 3](#page-21-0) below, PHP's actual load over the past six years (2020-2025) has also - been difficult to accurately forecast. Importantly, the data in Table 3 is forecast data from PHP, 78 7 the - 8 entity best...

AI summary The text discusses the difficulty in accurately forecasting PHP's load over the past six years (2020-2025), noting that PHP, being the entity best suited for forecasting, has experienced an average forecast error and a peak error in 2023.

Table 3. Forecast Error in PHP Annual Load (2020-2025)[79](#page-21-2) 11 Year PHP Forecast Load (GWh) PHP Actual Load (GWh) Forecast Error (GWh) Forecast Error (%) 2020 2021 2022 2023 2024 2025 Average Forecast Error 13 Q. Does PHP agree...

AI summary The document presents a table showing forecast errors in PHP's annual load from 2020 to 2025 and asks whether PHP agrees that it has difficulty predicting its own load.

- 14 A. Yes. PHP noted that "[f]orecasting the required amount of annual energy for more than one year - in advance is challenging due to the nature of the paper market."80 15 Consider, too, that PHP indicates that - its forecasted load fo...

AI summary PHP acknowledges the difficulty in forecasting annual energy requirements more than a year in advance due to the paper market's nature. It also notes a 5% reduction in its 2026 forecast compared to the one provided to NSPI for the General Rate Application proceeding.

CONFIDENTIAL (Attachment Only) 1 (iii) ADC load shifting benefits. 2 3 (h) Please compare the magnitude and causes of forecast error in 2025 with those in 2024 4 and prior years. 5 6 Response IR-5: 7 8 9 Preamble: The reference provided in...

AI summary The document discusses discrepancies in load forecasts for 2025, including the incorrect quotation of a demand requirement and the sources of different forecast figures. It outlines the ELIADC Tariff process and how revised load forecasts are handled by NS Power and PHP.

Reference: N-20, Bates White Evidence, at p. 21, lines 2-6 notes a second concern: Our second concern is that the DR credit is built upon a shaky premise, namely, that PHP's load is sufficiently predictable as to allow for an after-the-fac...

AI summary Bates White Evidence raises concerns about the reliability of PHP's load forecasting for demand response (DR) credits, citing volatility, forecast errors (averaging xxxx/year, peaking at xxx GWh in 2023), and external variables (e.g., equipment performance, macroeconomic factors) that complicate accurate cost-based load (CBL) calculations. It questions NSPI's ability to model CBL without PHP's operational data and highlights risks of cost variability tied to forecast inaccuracies.

el to better reflect likely PHP net load, capturing both updates to expected PHP gross load and, most importantly, the likely output from Goose Harbour at a facility capacity of 168 MW." Do you agree. A. While we currently only have SIS ap...

AI summary The discussion centers on the use of the P90 value for wind power projects, specifically the Goose Harbour facility at 168 MW, and whether a true-up mechanism is needed for PHP net load deviations. The response disputes Bates White's recommendation, citing corrected assumptions about the facility's output and usage.

- 31 (f) What annual capacity factor does Bates White believe is more reasonable 32 than the noted 44% annual capacity factor? Please provide the studies or 33 comparables relied upon in answering. 1 (g) In principle, is it appropriate to...

AI summary The document includes questions regarding the annual capacity factor, use of P90 or P50 estimates, and concerns about the predictability of PHP's load and forecast errors. Bates White raises issues about the reliability of assumptions used in setting the ELID Energy Charge and the challenges in forecasting PHP's load due to various operational and external factors.

(a) Please quantify the term "substantial" in this context. (b) Please confirm that this true-up mechanism is meant to operate in both directions and could provide a credit per MWh in favour of PHP if the actual PHP net load is higher than...

AI summary The text includes three questions aimed at clarifying the meaning of 'substantial,' the functionality of a true-up mechanism in relation to PHP, and the methodology NSP uses to forecast energy usage for LIIR customers.