=== dc2018-04-0008-pages-01.ppm === Republic of the Philippines DEPARTMENT OF ENERGY DEPARTMENT CIRCULAR NO. PC201R- €4- Coos ADOPTING FURTHER AMENDMENTS TO THE WHOLESALE ELECTRICITY SPOT MARKET (WESM) MARKET MANUALS ON BILLING AND SETTLEMENT AND LOAD FORECASTING METHODOLOGY FOR THE IMPLEMENTATION OF ENHANCEMENTS TO WESM DESIGN AND OPERATIONS WHEREAS, Sections 30 and 37(f) of tne Electric Power Industry Reform Act (EPIRA) provides that the DOE. jointly with the electric power industry participants, shall establish the Wholesale Electicity Spot Market (VWVESM) and formulate the detailed rules governing the operations thereof: WHEREAS, on 28 June 2002, the DOE, with the endorsement of the electric power industry participants. promulgated the WESM Rules through Department Circular No. OG2002-06-003; WHEREAS, any changes, amendments, and modifications to the WESM Rules including its Market Manuals shall be undertaken in accordance with the provisions of Chapter8 thereot. WHEREAS. on OF December 2015, the Philippine Electricity Market Corporation (PEM) submitted and presented tc the Rules Change Commitiee (RCC) the Proposed Amendments to the WESM Market Manuals on Gilling anc Settlement (BS) and Load Forecasting Methodology (LFM) for the implemeniation of enhancements to WESM design and operations WHEREAS. the PEMC's proposed amendments to the said Market Manuals aim to facilitate the implementation of the new market features as provided under the DOE policies on enhancements to market design and operations and the amended WESK Rules for the New Market Management System (NMMS) through Department Circulars No. DG2015-10-0015 and No, 0C2016-10-0014, respectively; WHEREAS, on 16 December 2016, the RCC approved the publication of the said PEMC's proposal in the WESM website to solicit comments from market participant and other interested parties. WHEREAS, on 03 and 17 March 2017, the RCC during its 126" and 127" RCC Meatings respectively, deliberated on the said proposal giving due course to the somments recerved from market participants and the DOE; WHEREAS, on 11 April 2017, the RCC during its 128" RCC Meeting, further deliberated on the proposal taking into consideration the DOE Department Circular No. BC2017-03-00027 dated 20 March 2017 concerning implementation of Must Energy Canter, Rial Grove. Goreiecio GrogelSoy. Taguag Gir, Mager Mania, Prilippines fT Tate: (Troms) S4-1407 we 4d): Fi B02 B02: 8-2 067 (Dometic): aee-2 17S: A Gee 208 linthernatioreal| === dc2018-04-0008-pages-02.ppm === Dispatch and Priority Dispatch Generating Units in the WESM, and it finalized and approved the proposal for endorsement to the PEM Board: WHEREAS, on 18 September 2077, after due evaluation and deliberation, the PEM Board during its 127" PEM Board Meeting approved for endorsement to the DOE the above stated RCC proposal: WHEREAS, on 02 October 2017, the PEM Board-approved amendments to the WESM Market Manuals on BS and LFM were submitted to the DOE for fina! approval, in compliance with Chapter 8 of the WESM Rules: WHEREAS, the DOE reviewed the sad PEM Board-approved proposal, made minor revisions for consistency and deamed it consistent with the objectives of the WESM and the DOE policies on the enhancements of WESM design and operations: NOW THEREFORE, pursuant to its authority under the EPIRA and the WESM Rules, the DOE hereby adopts, issues, and promulgates the following amendments to the WESM Market Manuals for the implementation of enhancements to WESM design and operations: Section 1. Amendments to the WESM Market Manual on Billing and Settlement. The following provisions in the Billing and Settlement Manual are hereby amended: (a) Section 1.1 under Introduction is amended to read as - “1.1 Background This Market Manual provides the procedures on the preparation of settlement statements as basis for the settiement of transactions of Trading Participants in the Wholesale Electricity Spot Market (\WESM)." ib) Section 1.2 under Introduction is arnended to read as - “1,2 Purpose This Market Manual implements relevant provisions from WESM Rules Clauses 3.74 and 3.18. lt deseribes the procedures which the Market Operator and Trading Participants must follow in relation to the settiement process and compliance with the prudential requirements.” ic} Section 1.3 and its subsections under Introduction are amended to read as - "1.3 Scope 1.3.1 This Market Manual covers tne following procedures: a) Section 4 - Issuance of Settlement Statements in accordance with WESM Rules Clauses 2.74.4, 3.14.5, 3.14.8, and 3.14.9: Saas oo = Adophng Eure: Amentenente fh Ene Waele Ele carhcity Spo Merce: (WWESM) Markel Marais on Billing and Settamant and Loac Féerscasting Methotalogy far the bnplemenighen of Enhasoerniens to WESM Design and Operations 2/12 === dc2018-04-0008-pages-03.ppm === b) Section 5- Collection and Payment of Setiement Amounts in accordance with WESM Rules Clauses 3.14.2, 3.14.3, 3.14.6, 3.14.7, and 3.14.10; ¢) Section 6 - Payment Default in accordance with WESM Rules Clauses 3.14.11 and 3.14, 12; d) Section 7 - Prudential Requirement in accordance with WESM Rules Glause 3.15; and 2) Section 6 - Suspension and Revocation of Suspension in accordance with WESM Rules Clauses 3,74.11.2(b) and 3.15.8. f) Section 9 - Bilateral contract declaration for energy and reserves in accordance with WESM Rules Clauses 3.13.1 and 3.13.2; and g) Section 10 - Filing of additional compensation in accordance with Secton 6.3 of the Price Determination Methodology, 1.32 This Market Manual does noi cover the following processes, which are ali provided under the Price Determination Methodology Manual: 8) Calculation of administered prices: ob) Settlement amounts calculation; c}) Net settlament surplus: and d) Must-run unit catculation.” (d) Section 2.1 and its subsections under Definitions, References and Interpretation are amended to read as - “2.1 Definitions 2.1.1 Unless otherwise defined or the context implies otherwise, the italicized terms used in this Market Manual shall bear the same meaning as defined in the WESM Rules and other Market Manuals. 2.1.2 The following words and phrases as used in this Market Manual shall have the folowing meaning: a) Average Actual Market Price. This refers to the ratio of the total spot market payment of a WESM Member, wnich may include spot market energy and reserve transactions to the tote! metered quantities net of bilateral contract quantities for each billing month. b) Credit Support Provider. Banks, or insurance companies. c) FIT-4ll Administrator, The Wational Transmission Corporation as administrator of the FIT-All Fund, as designated in Resolution No. 15, Series of 2072, amending the FIT Rules. da) Marginal cost of congestion. Congestion component of the locational marginal price as defined under Section 4.10.1 of the Ponce Determinaton Methodology. &) Marginal cost of losses. Losses component of the locational marginal pnce as defined under Section 4.10.9 of the Price Determination Methodology. fi Preliminary Statement A statement issued by the Market Operator under WESM Rules Ciause 3.14.4. —EEEE saan eeeeeeeeeeeneeeeeeeeeeeeeeeeeee [ae = Agepting Further Amancments to the Wholesale Elecinoty Spot Market (ESA) Markel Manusis an Billing and Setiement and Load Forecasting Melhodology for ihe implementation of Enmancements to WESH Gen anc Operations ae === dc2018-04-0008-pages-04.ppm === g) Setiement. The activity of producing bills and credit notes for WESM memSers in accordance with WESM Rules Clause 3.13, and with the processes defined in WESM Rules Clause 3,14 h}) setlement Statements. Statements issued by the Market Operator that contains the settlement transactions of a WESM Member. It can either be 3 preliminary, final or revised statements.” (@) New Section 2.2 under Definitions, References and Interpretation is added to read as = “2.2 References This Market Manual shall be read in association with the WESM Rules and other relevant Market Manuals.” (f} New Sections 2.3, 2.31 and 2.3.2 under Definitions, References and Interpretation are added to read as - “2.0 Interoretation 2.3.1 Any reference to a clause in any section of this Market Manual shall refer to the particular clause of the same section In which the reference is made, unless otherwise specified or the context provides otherwise. 2.3.2 Standards and policies appended to, or referenced in. this Market Manual shall provide a supporting framework.” (9) Section 3.1.1 under Market Operator is amended to read as - “3.1.1 The Market Operator shall be responsible for complying with the requirements set forth in this Market Manual and in the WESM Rules, as follows: MXE MEX uN" (h) Section 3.2 under Responsibilities ls amended to read as - “d.2 Trading Participants The Trading Participants shall ba responsible for complying with the requirements set forth in this Market Manual and in the WESM Rules, as follows: a} Retrieve and review settlement statement files and supporting data issued by the Market Operator. The Trading Participants shall notify the Market Operator if files are not accessible or received within the timetable and if there aré discrepancies or errors; Qn ERX xt KEM Pe — Adoming Further Amengment: to the Wihobasals Elactooty Spal Maret (WESM) Marke! Manuva on Bling and Setlamantand Load Forecasting Méihadology for the bnpementajon of Enhanoemerits 4c ]2 to WES Design anc Oppereiticts === dc2018-04-0008-pages-05.ppm === (i) Section 4.1.1 under Contents of Settlement Staternents and Data j= amended to read as - “4.1.1 Settlement Quantity and Amounts a} The settlement quantity billed to each Trading Participant is in accordance to WESM Rules Clauses 3.13.5 and 3.13.6. 6) The trading amount billed to each Trading Participant is in accordance to WESM Rules Clauses 3.13.17 and 3.13.18. c) The reserve cost recovery charge billed to each Trading Participant is in accordance to WESM Rules Clause 3.13.9, a) The transmission rights trading amount billed to each Trading Participant is in accordance to WESM Rules Clause 3.13.70. ©) The aggregate trading amount for a Trading Participant for a settlament Interval equals the sum of: . The energy trading amounts for each market trading node for which the Trading Participants is responsible calculated in accordance with WESM Rules Clause 3.13.7 (which will typically be positive for a Generation Company and negative for a Customer): plus i. The reserve trading amounts for each reserve region into which that Trading Participant contributes reserve calculated in accordance with WESM Rules Clause 3.13.8 (which will always be positive for both Generation Companies and Customers); plus ii. The transmission night trading amounts for each transmission right held by the WESM Participant calculated in accordance with WESM Rules Clause 3.13.10 (which will always be positive for both Generation Companies and Customers) olus iv. The reserve cost recovery charge determined for that Trading Participant with respect fo any reserve cost recovery zone within which it has any facility connected calculated in accordance with the procedures developed under WESM Rules Clause 3.3.5 (which will be positive for any Trading Participant}; and v. Any other ancillary service cost recovery charges determined for that Trading Participant in accordance with the procedures developed under WESM Rules Clause 3.3.5. ft) For each billing period, the Market Operator shall determine the settlement amount for each Trading Paricipant as the sum of the aggregate trading amounts for the settlement intervals in that billing period, determined in accordance with WESM Rules 3.173.112 plus: |, Any amount payable by the Market Operator to that Trading Participant in respect of that billing period and not accounted for in WESM Rules 3.13.11.2, including payment for any ancillary servicas purchased on behall of the System operator, bess the sum of li. Any market fees which that Trading Participant is required to pay in respect of that bliling period as determined im accordance with WESM Rules 2.10; plus Dt. — AdopnG Furher Amendments i ihe Wholesale Elecinoty Spot Marcel (WES) Market Manuals on Bling and Settlement ane baad Forecasting Meatnodoiogy tor the implenentation of Enhanoaniant io WESM Design and Operstions a afiz === dc2018-04-0008-pages-06.ppm === ii. Any other amounts payable by that Trading Participant to the Market Operator in respect of that billing period, including any reserves recovery charges.” (j) Section 4.1.2 under Contents of Settlement Statements and Data is amended to read oe - “4.1.2 Net Settlement Surplus. if the transactons indicated in Section 4.1.1, in aggregate, results in a surplus or deficit ramaining, this will be known as the ret settlement surplus. This may be retained by the Market Operator, to fund deficit as a resull of transactions required in WESM Rules Clause 3.73.11 or may be flowed back to the Market Participants in accordance with the procedures to be developed under WESM Rules Clause 3,13.12.3 or may be used by the Market Operater to establish and support the market for Financial Transmission Rights subject to (he approval of the PEM Board,” (Kk) Onginal Section 4.1.3 (Price Substitution Adjustments) under Contents of Settement Statements and Data is hereby deleted. (1) Orginal Section 4.1.4 (Must-Run Adjustment) under Contents of Settlement Statements and Data is renumbered and amended io read as - “4.1.3 Must-Run Adjustment. in a particular billing period where a must-run generating unit was implemented for a dispatch inierval, the settlement for the Trading Participant of the generating unit for the affected dispatch interval shall be in accordance to the relevant market price under the Price Determination Methodology Manual and Section10 of this Market Manual.” (m) Orginal Section 4.1.5 under Contents of Settlement Statements and Data is ranumbered and amended to raad as - “4.1.4 The settlement data that are being transmitted to the Trading Paficipants aiong with the preliminary and final siatemenis inciuding the market fee statements which includes the energy trading amounts, final nodal energy dispatch prices, energy settlement quantity, line rental trading amount of participants with bilateral contract quantity and each of the participants’ bilateral contract quantity, reserve trading amount, reserve recovery amount, net settiement surplus rebated, and line loss anc congestion charges. These set of information are being transmitted to the Trading Participants vie email and CD copy mailed along with the Participants Statements.” in} New Section 9 (Bilateral Contract Deciaration) and its subsections are hereby added to read as - “9 Bilateral Contract Declaration 9.1 Deciaration for Energy Transactions 6.1.1 Trading Participants who will sell electricity pursuant to bilateral contracts during @ dispatch interval and wish those bilateral contracts to be be = Adopting Further Amendments io the Wholesale Eustirioty Spot Market TWESM) Mercat Manueis & Billing and Seatiement and Laad Forecasting Mainodology tor tne bmolementation of Enhancements to WESM Daesion and Operasans Hy12 === dc2018-04-0008-pages-07.ppm === accounted for in settlements may declare to the Market Operator bilateral contract data up to the end of the following trading day. 8.1.2 To be valid, bilateral contract declarations submitted by a Trading Participant shall include: @. Market trading node from the identified market trading nodes during the enrolment of the supply contract designated as the source of the supply from the selling Trading Participant: 6. the Trading Participant woo will buy the declared volume pursuant to their Hilateral contract; c. market trading node fram the identified market trading nodes during the enrolment of the supply contract and whose final nodal energy dispatch price will be used as reference during settlements: and d. bilateral contract quantity in MVVh for each dispatch interval for that tracing day. 0.1.3 Bilateral contract dectaratians may include the market trading node from the identified market trading nodes during the enrolment of the supply contract designated as the withdrawal point of the buying Trading Participant. 9.1.4 Bilateral contract declarations submitted by a Trading Participant should be covered by a supply contract enrolled by the Trading Participant to the Market Operator In accordance with the relevant Market Manual, 9.1.5 The Market Operator shall immediately send to each Trading Panicipant with whom it has received a valid bilateral contract declaration an electronic confirmation of receipt af that bilateral contract declaration. 8.1.6 lf the enrollment of the supply contract covenng the bilateral contract declaration requires confirmation from the buying Trading Participant, the valid bilateral contract declaration shall be accounted for during settlements niy upon confirmation by the Buying Trading Participant in accordance with Section 9.2. 8.1.7 lf the enroiiment of the supply contract covering the bilateral contract declaration does not require confirmation from the buying Trading Participant, the valid bilateral contract declaration shall be accounted for during settlements unless nullified by the Suying Trading Participant in accordance with Section 9.2, 9.1.8 If the bilateral contract deciaration is invalid, the. Market Operator shall promotly inform the Trading Participant and such bilateral contract declaration shall not be accounted for in settlements. 9.2 Confirmation and Nulllification for Energy Transactions 9.2.1 In order to be accounted for during settlements, valid bilateral contract deciarations covered by supply contracts enrolled to require confirmation submitted by a Trading Participant should be confinned by the buying Trading a Be = Jibopbng Further Sriendmets th the Wholesale Elactiony Spot Market PVWESM) Market Manuals on Billing and Settiement and Load Forecasting Methodology for ihe impiementaion of Enhancements to WESM Design and Operation = i f === dc2018-04-0008-pages-08.ppm === Participant to the Market Operater one day after each trading day. 9.2.2 If the enroliment of the supply contract dogs not require confinnation from the buying Trading Participant, the valid bilateral contract declarations submitted by a Trading Participant shall be accounted for during settlaments unless the buying Tracing Participant submits to the Market Operator a notice of nullification one day after each trading day. 9.2.3 The Market Operator shall immediately send to each buying Trading Participant with whom it has received a confirmation or notice of nullification an electronic confinnation of raceipt of that confinmaiion or notice of nullification §.2.4 Upon receipt of a confirmation or notice of nullification, the Market Operstor shall immediately notify the relevant selling Trading Participant to the supply contract that the valid bilateral contract declaration has been confirmed or nulliied. 9.2.5 At the end of each trading day. the Market Operator shall notify ail buying and selling Trading Participants, whose supply contract was enrolled io require confirmation and whose valid bilateral contract declarations was noi confirmed, that their bilateral contract declarations shall not be accounted for during settements. 9.2.6 Al the end of each trading day, the Market Operator shall notify all buying and selling Trading Partiapants, whose supply contract was enrolled to not require confirmation and whose valid bilateral contract declarations was not nullifiad. that their bilateral contract deciarations shall be accounted for during settiaments. 9.3. Line Rental Calculation for Energy Transactions Gd.) WESM Rules Giause 3.13.7.2 provides that the Market Operator shall calculate line rental trading amounts for each bilateral contract quantity declaration, for informational purposes, 8.3.2 For the purpose of calculating line rental trading amounts, bilateral contract quantities deciared for buying trading participants shall be: a. if available. assigned to each of its market trading nodes as identified in Section 9.1.3; or b. pro-rated to each of its market trading node basec on metered quantities. 6.3.9 For a selling Trading Participant, the line rental trading amaouni associated with @ bilateral contract declaration shall be equal to the bilateral contract quantity multiplied by the differance of the final energy dispatch price at the market trading mode identified under Section 9.1.2{a) and the final energy dispatch price at the market trading node identified under Section 7 2c. I ee eee ee ee Ee oe ee Do. — Adopling Further Amendments fo the Wiholeseee Electiaty Gaol Market (WWESM) Marset Maruais on Billing and Setthereent ond Load Forecasting Mesthodology for the bnstementabon of Erhancenients 19 WESM Design and Operations Bi? === dc2018-04-0008-pages-09.ppm === 9.3.4 For @ buying Trading Participant. the line rental trading amount associated with a bilateral contract declaration shall be equal to the sum of the line rental trading amounts calculated for each of its market trading nodes. The line réntal trading amount for each market trading node of a buying Trading Participant is equal to the bilateral contract quantity or, if applicable, the pro-rated bilateral contract quantity calculated under Section 9.4.3 of the Markel trading nodes multiplied by the cifference of the final anergy dispatch price al the market trading node of the buying Trading Participant and the final energy dispatch price al fhe market trading node identified under Section 9,1.2{c). 9.3.5 The line rental trading amount associated with each bilateral contract declaration of a Trading Participant may be segregated inion: a. the line rental trading amount attributable to losses calculated by applying the same formula in Sections 10.1-3 and 10.7.4 but only using the marginal cost of losses component of the final energy dispatch prices in piace of the final energy dispatch prices: and b. the line rental trading amount attributable to congestion calculated by applying the same formula in Section 10.1.3 and 10.1.4 but only using the marginal cost of congestion component of the final energy dispatch prices in place of the tinal energy dispatch prices. 9.3.6 Line rental trading amounts segregated into ihe amounts attributable to the cost of lossas and of congestion shall be calculated by the Markat Operator and included in the sefilement statements of Trading Participants. Additional information to validate segregated line rental trading amounts may be provided by the Market Operator subject to the pertinent provisions of the WESM Rules and the relevant Market Manual. 9.4 Declaration for Reserve Transactions 9.4.1 The System Operator shall submit to the Market Operator the following information on bilateral contracts for reserves up to the end of the following trading day: a. Trading Participant who provides ihe ancillary service: 6, Reserve bilateral contract quantity; and c. Reserve category. &.4.2 The Market Operator shall confirm the receipt of reserve bilateral contract quantity declaration,” (o) New Section 10 (Filing of Claims for Additional Compensation) and its Sub- sections are hereby added to read as - “40 Filing of Claims for Additional Compensation 10.1. Criteria for Additional Compensation me pc - Adopting Further Amendments to the Wholesale Eiecticty Soot Manont (WESM) Markel Manuals on Billing and Sothemant ance Load Forecasting Metiocoiogy for the Implementation of Enhancements to WESM Deaton and Operators aie === dc2018-04-0008-pages-10.ppm === In accordance with Section 6.3.3 of the Price Determination Methodology Manual, Trading Participants may be entitled to additional compensation when the cost incurred in complying with the dispatch instruction are not sufficiently covered by the trading arnounts related to settlement intervals with dispetch intervals under any of the following conditions: a. Market suspension or Market intervention: or b. When a trading parbcipant is designated as must-run unit of comstrain-on generating unit. 10.2. Filing of Claims 10.2.1 Trading Parti@pants shall submit the written claim for additional compensation within the allowable timeframe, as follows: a. Marke! suspension or market intervention — fourteen (14) working days afier ihe resumption of the market and b. Must-run unit or constrain-on generating unit — within one (1) year after the trading participant was designated as MRU or constrain-on generating unit. Any claims not filed within such period shall ba daamed waived. 10.2.2 Trading Participants shall submit sufficient proof regarding the costs incurred, which are limited to fuel cost and vanable operating and maintenance costs, which may include start-up and shut down cost. Below is the non-exhaustive list of raquirament documents in filing claims for additional compensation: a. Certified correct Fuel Consumption and Inventory Report; b. Purchase Invoices, Official Receipts and other supporting documents; and c. ERG approved rate or List of Vanable Operation and Maintenance Costs supporied by photocopies of invoires/receipts. 10.2.3 The Market Operator shall determine validity of the costs incurred based on the aforementioned supporting documents. 10.2.4 The Market Operator shall inform the requesting Trading Participant of the approval or disapproval of the claim within fourteen (14) working days from receipt of the complete documents from the Trading Participant. Any claim not decided within fourteen (14) working days shall be deemed Bporoved and shall be allocated and billed immediately in tne succeeding billing period.” (p) Original Section @ (Amendments, Publication and Effectivity) and its subsections are renumbered accordingly to read as — “44 Amendments, Publication and Effectivity 114 Amendments oe — Adopting Further Amendments to the Wholessie Elecincity Spat Market WES) Maeriogt Manuees on Billing and Setlement and Load Forecasting Mathocoogy tor ine lmaementaton o Enhancements to WWESM Desegn and Operations 1/1 === dc2018-04-0008-pages-11.ppm === XXK | XXX XR 11.2 Publication and Effectivity MX KX eax (9) New Appendix “A” is hereby added to read as - cee em ee = PROCEDURES TIMETABLE | Declaration of Bilateral | Within one (1) day following the trading day | Contract Quantities for | Energy and Reserves ‘Filing of Claims for|a. Market suspension or market intervention | | Additional — fourteen (14) working days two billing Campensation periods after the resumption of the market: and b, Must-run unit or constrain-on generating unit — within ome (1) year after the trading participant was designated as MRU or | constrain-on Generating unit Section 2. Amendments to the WESM Market Manual on Load Forecasting Methodology. The WESM Market Manual on Load Forecasting Methodology Issue No. 3.0 and presented as Anné= A ip this Circular is hereby approved and adopted. For this purpose, the Philippine Electricity Market Corporation (PEMC), as the Market Operator, shall expedite the development of the procedures in connection with the preparation of load distribution factor as required under Section 6.2.3 of the Load Forecasting Methodology Manual. Such procedures to be developed and approved in accordance with the WESM Rules shall form part of the said Markel Manual for consistency and transparency. Section 3. Transitory Provisions. These amended Market Manuals shall be iniBally weed for thé deveiooment of tha WESN''s New Market Management System (NMMS) and provide guidance for the stakeholders and VYESM panricipants on the implementation of the enhancements to WESM design and operations. For clarity, the current WESM Rules and its Market Manuals shall remain valid and effective for ihe commercal operation of ihe integrated WWESM and Retail Market using the present MMS, Section 4. Separability Clause. If for any reason, any section or provision of this Circular is declared unconstitutional or invalid, such pars not affected shall remain valid and subsisting. a a a a oc = Jgeping Further Amendmenis ft the VWholesaie Eleciioaty Spot Marcel (WESA) Marka Manuak.en Billing and Setiiement and Load Forecasting Methodology for the Impemaniation of Erhancements ia WESM Desegn and Operalons Maz === dc2018-04-0008-pages-12.ppm === Section 5. Effectivity. This Circular sha!l take effect upon its publication in at least two (2) newspapers of general circulation and shall remain in effect until otherwise reverkexd, lssuec this February 2014 at the DOE, Energy Center, Riza! Drive, Bonifacio Global City, Taguig City, Metro Manila. io y\ FoNs6'6. cust’ cretary Fae oe Poe BP a Te oe ee EDR Sees oe OOE-AGC- oes CP BY OO TRAN APA MAR 28 2018 oe — Adopting Further Amendment io in Wholesale Electnot, Spot Market TWESM) Market Marudi on Bing and Settement.and Loan Forecasting Matnodoiogy for tha Implementation of Enhantemens IO WESM Design and Operabons re === dc2018-04-0008-pages-13.ppm === Sp Wholesale Electricity Spot Market “Annex A" Load Forecasting Metreadology WESM-LFRM-03 PUBLIC WESM Market Manual Load Forecasting Methodology Issue No. 03 Absiract This Market Manual sets out the procedures for determining load foracast used ir the reigevant mamket runs in the Philippine Wholasale Blectincity Spo1 Markat Document denny WESK-LFMAGS lsaue. oa Reason for teas: Revigions in methodologies i view of amhancements to WESM desian and operations Effective Gate: lesue 03 -WESNM-LFIM Publis === dc2018-04-0008-pages-14.ppm === &>: Wholesale le Electricity Spot Marke Load Forecasting Methodology WESM-L FRIAS Document Change History | Issue No. Proponent eiactiony Reason for 4mendmert | 1.0 MO. Sub-com 2004-08-09 | New Document 0.1 WO Sub-com 2004-08-17 | Sub-com members comment incorporated 2.0 PEMWIC 2014-04-17 | Entire document revised in | accordance with recommendations | mace in the report an the operational | audit of the market operator systems | | and processes. 3.0 PEMC | Revised to incorporate changes to | forecasting methodologies and | definition of nodal load distribution | factors in accordance with the enhancements to WESM operations under DOE DC 2015-10-0015 Document Approval | RCC PEM Board DOE DOE Issue | RCC F PEM Board Na. Approval | iil Approval oe Approval ye seloiag on OF oi | AdO4-(4. nia nia 08 Jan 2014 204-02 05 Feb 2014 fa na | 2 17 Apr 2o17 | 2077-05 Reference Documents Decument ID. | Document Title | /WESM Rules | [ —~*SRilippin Gri oT | Digpaten Protocoi [WESM-UP | PA Consulting Group. Pniliogine Fiecticity Markel Corporation 2077 iIndenendent Aud? Summary Aeoart (12 Decamber 2012}, issue 09 -WVWESM-LFM Public i === dc2018-04-0008-pages-15.ppm === €! Wholesale Electricity Spot Market Load Forsomsting Methodology WESSEL PMAOS Table of Contents SECTION 1 INTRODUCTION ............0....-ccceceseeeenssecnssronsesnassnaussseaseurcesceneacscuneecenes 1 I cases ctccrine ccc pst vn ney ater aadSiT ichiek ev ecaenbiaebca bbs AaN NLR 1 Tt Pees asics ie eee ae ae Fl Tad, SOP EL. ssccesrsenpenssteteensesnnncesaseecossanceateeseeneseenpuseneeneeretneas sunstinasdéctieesacesene 3 SECTION 2 DEFINITIONS, REFERENCES, AND INTERPRETATION.................3 1, DEFINITIONS........ oui AY RAGE RMPL ESTE PEE CoAT RTE at ne a 2.2. FREFERENGES oo. .cccsssnneeve ess seeecnnasseeaeeceseeceeceaseeeeneseaseceraspesesnansnnanunenausauvuusemenees 4 2.3. INTERPRETATION iiistecssissssscsenes SN a ees - SECTION 3 RESPONSIBILITIES | ......).:00cc......cccc0cccceseeeseceesyeneeesensseneenssenesseusseneses 5 PRE NO occ so a norcacsnarnracy ior eceav seleearauers eres nahtaeraaeaLsaaateatas 5 doe. SVSTEM PERATOR Sucve a ae 5 3.3. TRADING PARTICIPANTS .,....::::cccsssessscesssensneeeeneaseeeeeseresscenetesteneseunnrrecwrnarenaes 5 SECTION 4 SHORT-TERM LOAD FORECAST ciccsiccccccccsccccccrscsisisnevtaearseavaeaes 2 1, BACKGROUND |... cetsveeeneers ESE OCP EAT INE ie PR ERLE rE Moa SSR ME aR SRS 5 4.2, DETERMINATION OF HOUALY DEMAND FORECASTS ooo ese ees 6 SECTIONS VERY SHORT-TERM LOAD FORECAST ..0.00000000. 000.02 ee T Ti BRAC oo cece cee cece cendecea ed 5 (dake p fade eka bash ak weengbdd Shei Seite eeeer eure ses f 5.2. DETEAMINATION OF DEMAND FORSCASTS FOR EACH DISPATCH INTERVAL........ 8 SECTION 6G NET LOAD FORECASTS. citi ce 5 Bi, SOOORE.. cevvccscccessveecaveeeeserenssrivrerrasesapecenagsessussasuvang pass auevesaiessurstssaeescaatesssaeneee Fs] 6.2. UNAESTRAINED NET LOAD FORECAST wiccccecciccccess csessetics -cteeicaetvssescnevenseredeberiey 8g 6.3. AESTAAWED NET LOAD FORECAST ii... cc.ccc elec ccc ec jggeeespeneetee. 10 6.4. INITIAL LOSS PERCENTAGE oo... cic cette sees ceecestececeescuserenusecenenennaaees 10 SECTION] CUSTOMER FORECASTS. ...0...-..ccccss:ccscecssreneess ieelitehvssdinrrseseiciea! 4 7.1. COSTOMER FORECAST SUBMISSION ..0.:...10ccstccssensteecssseescssssueasssuscssnsssoasioes 17 7.2. CUSTOMER LOAD FORECAST CONSIDERATIONS ....-.ccccaveresowssencusseaesreensnneass 12 SECTIONS LOAD FORECASTING PARAMETERS ....-.:ccccctcccccccsussereee sata CS re 12 Bie BARRA cc ceescicesercenn scene ceed ereidy sateen waa eine tanga enact at 13 B.3. POWERS VS TEM COMDITION #.iiicccjoecccceeee cesscessseetceabelusdbseescnscabiseuisessaniesaes 13 issue 09 -WESH-LFM Piiblic i === dc2018-04-0008-pages-16.ppm === ©: Wholesale He Electricity Spot Market Load Forecasting Methodology WESM-LFMA SECTION 1 INTRODUCTION 1.1. Background Lit. Load forecasts are necessary inputs fo the optimization runs in the Wholesale Glecticity Spo! Marke! (WESM), particularly the week.thead oraiection, day- 4oead oroection. Nour-ahead protection, and the real time disasich market runs, The responsibility to prepare fogo forecasts rests in the Marke! Gnerstor although Customers are also permitted to perform their own customer forecasts under certain conditions. 1.1.2, The WES Aules provide for the responsibilitias of the Marie: Ooersfor in penorming tS Joao forecasts: a. VES Aules Clause 3.7.4.1 (c) requires that each marke! Grojection shall take into account the forecast demand information prepared by the Market Onerator in accordance with bVESM Ales Clause 3.5.4, b. WES Rivles Clause 3.7.4.2 further requires that priar to the preparation of each set of marke! proections, the Martet Operstor shall, in consultation with the Sysiem Operator, prapare an expected unrestrained oe! joad forecast! in accordance with the procedures developed under Clause 3.5.4, and such number of other load scenanos as may be datermined in consultation with WES Panicijants and approved by the PEM Board c. For the rear-iime dispatch marker mus, WES Auies Clause 3.8.17 (b) and (c) directs that at thé beginning of each interval, the Marker Operator shall prepare a forecast of the wrested met oad expected al each market trading node for the end of that trading interval, and adjust that voresirained net igad forecast to account for joad shedding, if required, in accordance with V¥ESI Auies clause 3.9.5. d. Furthermore, “ESM! Sules Clause 3.5.4.2 states that each net iosc forecast! shall be prepared in such 4 Waly 45 to represent the net jose to be met by generation fram scheduled generating wos, musi-dispatch geneanaiing units, priarniy cispalch generating units, and mor-scheduise qeneraling unis including losses occurring outside the system represented by the marke’ network model, but excluding any SEneouwied faag, issue 2.0 -WESM-LFM Pubic + === dc2018-04-0008-pages-17.ppm === Wholesale Electri } Spot Market ae Load Forecasting Methodology WESHLL FM-O3 1.1.3... 1.2. eB, WESM Rules Clause 3.5.4.3 states that the unrestrained ne! joad forecast! shall be prepared $0 as fo represent the ne! load as it would be, or would have been, in the absence of (oad shedding. f. VESM Rules Clause 3.5.4.4 states that A load shedding is expected to OSU In any oospaeich inteval, a restrained nel iced forecast! for that gispatcn jiterva! shall be prepared on the same basis, bul accounting for joao shecoueg to the extent that itis expected to occur. “Lsfomers area permitted under “VES Aives Clause 3.5.4.1 to submit their own Joed forecast! that shall be used by the Martel Opensor in the preparation of net foal forecast! subject to certain conditions. a. Each Cusfomer may submit a forecast in respect of each olspaich interval for each of its registered /oad facilities for each trading day of week in accordance with the WES fimetabie. The fimeteble is set out in Section 4 of the bVWESiW Dispatch Protocal. b The forecasted load Misi be used by the Marke! Qoerafor in the preparation of the net oad forecest ea FF the Customer fais to submit @ forecast for his Joead facility. in accordance with the VES timetable then the forecast prepared by the iderke! Ooerator at ine relevant node shall be used. Purpose This Warke anual asiablisnas the following: 12.1, 1.22. The requirements in determining the /oad forecasis for thé following mane rons in the WES a. Marker Profactions L Weer-Ahead Proection (WAP) ii. Dey-Ahead Proectan (BAP) iii its Ahgad Projection {HAP) b. Realiirne Dispatch The forecast methodology to be employed by the MWarker Ooerafor for preparing load forecasts. issue 2.0 —\WESKELFI Public Zz === dc2018-04-0008-pages-18.ppm === p> Wholesale Electricity Spot Market Load Forecasting Matiodoiary WESM-LPML-OS 1.4. Scope The systems, processes and procedures set out in this Wanket \/anua/ shall be used In the preparation of ioad forecasts to be used for the weesx-atead oroectons WAP). Gayahead onmections (DAP), Nour-ahesd orajections (HAP), and real-time cSpalch marke! runsin the bvesi SECTION 2 DEFINITIONS, REFERENCES, AND INTERPRETATION 2.1. Definitions 2.1.1. Unless otherwise defined in this section or the context implies otherwise, the italicized terms used in this Marke! Manue! shall bear the same meaning as defined in the WESIY Aules, 2.1.2. In addition, the following words and phrases as used in this /werker Manus! shall have the following meaning - a. Demand. Total power consumed in a cower sysfam or god inclusive of the relevant losses produced during the delivery of power. It is also the total power generated in the ned b. Demand Forecast. O27s od projection for a particular /oreceas! area c. Final Demand Forecast. Sum of the ne! joo forecasts plus the bosses fie (OeceS? ares qd. Forecast Area. A aiid in the Philippine power system where Gemand foreces! will be applied. Currently, gach onc in the Philippines is designated as a /oreces! ares, namely the Luzon, the Visayas, and the Mindanaa ones B. Input Gemand Forecast, initially assumed camend forecast that shall be used as tne off-take for determining thé wiresirained ne! ised i Load Distribution Factor (LDF). Factor used in allocating the total projected joad to individual ioacs in the marke! nefwork model. jseue 2.0 -WESM-LFM === dc2018-04-0008-pages-19.ppm === >! CPS Wholesale e Electricity Spot Market Load Pgrecasling Meinadology WIESMLLFEM-O2 g. Market Management System (MMS). Infrastructure that supports the operations of the WES and which includes tunctionalities that support the processes set out in this Mente! Manual A. Real-time Dispatch, Otherwise known as ATD. It is the Gisoaich Schedule which determines the target loading of tacilities at the end of the dispatch interval, The RTD calculates the ex-ante nodal energy Cepatch orices. i Short-Term Load Forecast. Hourly demanc forecesis starting at the next hour up to seven days ahead, fl shall be epptied to the calculation of the week-ahead provection and day-shesc projectior j. Unrestrained net load forecast. Also refers to unrestrained load ki. Very Short-Term Load Forecast. Semand forecasts for each dispatch interval starting at the next cisparch inferya/ up to the next two hours. It shall be ba nia the calculation of the four-aheéad oroectian and PSal-fiihe Geshhalle l, Weather Area, A specific location where there is availapla weather data. 22. References. This Werke! late! shall be read In association with the WES Rives and other relevant Wanker Marvell 2.3. Interpretation 20.7. Any felerence to a clause in- any Section of this iverke! Manuva! shall refer to the particular clause of the same section in which the reference is made, unless otherwise specified or the context provides otherwise. 2.3.2. Where there isa discrepancy of conflict between this Marke! Viaenus! and the WESM Aules, the WES Aules shall prevail. 2.3.4. Standards. and policies. appended to, or referenced in, this: Marke! Manual shall provide a supporting framework, lseuie 2.5 -WESH-LFM Pubic a === dc2018-04-0008-pages-20.ppm === ‘$p) phatesals Electricity Spot Market Load Forecasting bMothoosoagy WESN-LFM-09 SECTION 3 RESPONSIBILITIES 2.1, Market Operator 3.1.1. The Marke! Operator shall prepare et iead forecasts for each customer marker tracing node in accordance with the reattiodedagy sei out in this iWiathe? ian ual a.1.2. The ‘Warker Operator shail be primarily responsible for the davelooment and review of the fosd Jorecasting methodology documenied in this Market MiaiTial a.2. System Operator 2.2.1, The System Opersior shall provide the information réquired by the Market QOoerator in aaccerdance with this. Maret Manual, and ensure such information's accuracy, a2.2. The System Operator shall assist the vane: Oparstor in improving the icac hrenasiag meihodology set out in tus. dancer ienval 3.3, Trading Participants 4.4.1. The Tracing Pacicipans shall carry out the responsibilities provided for in this Markel iMfarual a.3.2, Customer Trading Partcioants that opted to provide nal iosd forecasts for their respective facilities shall be responsible for the timely submission and accuracy of their forecast information. SECTION 4 SHORT-TERM LOAD FORECAST 4.1. Background 4.1.1. Hourly damand forecast: for gach forecast area from the ‘current day to the next seven (7} diac Shall be determined by the /farvel Cperafor using the Market Management System's Shond-Tern Load Forecast /STLF) application, . a jesus 2.0 -\WESHELFEN Public === dc2018-04-0008-pages-21.ppm === >! Wholesale Electricity Spot Market Load Forecasiing Medhedolony WESML FMA 4.1.2, 4.1.4, 4.2, 4.2.1. 4.2.2. 4.2.3. Pree Wai rae Me perege Fie Wretrene Cana The aforementioned hourly cermand forecasts shall be used as inpu! Semen forecasts to obtain the hOuTy unresiained net load forecasts for the ek-anesd projection (WAP) and the day-ahead projection (DAP). The summation of the oe! /oac forecast plus the losses computed for a foreres! rea alter the cilinaization pees wi the relevant markel rh fie, WAP or DAP) constitutes the fina! doriand farecas! for thal forecast anes. The procedures for obtaining the nef oad forecests from the demand forecasts using acd cisinbuten factors (LOF) are discussed further in Section 6.of this Market Manuai Determination of Hourly Demand Forecasts The process of determining hourty demand forecasts involve various input date such @s historical Gemanc, historical weather measurements, and weather forecast data. Each /ofecas! ares is mapped to a VVeathor Area, which has its own distinct historical weather and forecasted weather information. Houny cemand forecas!s are determined using the MMS’ STLF application's weather adaptive, similar day, and pattam matching algorithms as shown In the illustration below, The methodology behind these forecasting algorithms are further discussed in Section 11. | iagoranirng, | + Wadia Saha git ew * Sie Ce | + Eire Aabiits hirag tote Gero _ Sauey Cararst = : put Cacend Toren Hes artat. 3 Fore Leta Afra | Amacai? Btls CHAT T Figure 1. Process Overview for Determining Hourly Demand Forecasts via MMS-STLF 4.2.4. The Varket Ooernsior may use other forecasting algorithms to determme hourly Gemand forecasts for each forecast area, logue 2.0 -V'ESM-LFM Publec a === dc2018-04-0008-pages-22.ppm === Wholesale Electricity Spot Market Load Forecasting Malhodeogy WESALLERLSS 4.2.5. The demand forecasi determined by the Marker Operator outside of the MEMS’ forecasting algorithm shall be treated as an axternal demand forecas 4.2.6. If an extemal demand forecas! is provided by the Marker Operator in a relevant hourly interval, then the external demand foreces! shall be Considered as the inout demand forecast ragardiess of any existing ¢demanc arecest determined by the MMS. These hourly cemand forecasts shail be Included in the MO's performance standards, 42.7. Pursuant to vo sii Ries peat 3.7.4.2 and 3.7.4.3, the Marke! Ooerstar Shall prepare a Marke! orgecion corresponding to each fosd scenario as may be determined in consultation with WES Particisants and approved by the PEM Boars 4.2.8, Load scenarios shall be provided for each day-shead projection using the following sensitivities (increments/decraments) with respect to tha jou! demand forecasi provided: Table 1. DAP Load Scenarios DAP Load Scenario Increme Load Scenario 1 Load Scenario 2 -3% Load Scenario 3 op FBG Load Scenario 4 l +555 SECTION 5 VERY SHORT-TERM LOAD FORECAST 5.1. Background 6.1.1, Cemand forecasts for each S-minute interval in ne next two (2) hours in each forecsst ; 7 shall be determined by the 2 Operator using the Marker Managemen! System's Vary Shar erin Load Forecast (VSTLF) application. 5.1.2. The demand forecests for each S-minute interval in the next two (2) hours shall ba used as ‘nou! demand forecasts to obtain the unrestrained net load forecasts = for the Hour-shead oovection (HAP) and the real-time dispaich East lssue 2.0 -\VWESM-LFM Public rd === dc2018-04-0008-pages-23.ppm === >! aca pel pSeciny Spot Market Laad Forecasting Wethodalagy WESM-LFMAGG 5.1.3. The summation of the ne! /osd forecast plus the losses computed for a foreces! area alter the optimization process in the relevant HAP or RTD run constitutes the fina) demand foracast for that foroceast sree. 5.2. Determination of Demand Forecasts for Each Dispatch Interval 5.2.1, The VSTLF uses historical demane data and shall be represented in a time- series. S.ce, From the demand time-series data, demand forecast shall be determined using the Cubic-Spline Interpolation Method, which details are discussed in Appendix D, | SECTION &6 NET LOAD FORECASTS | 6.1. Seope Pursuant to VVESM Fuies Glause 3.5.4.1, this section provides the \arke! Opersiars preparation ae ne! ioad forecasts, which are either unrestrained or restrained. for merket prolections and real-time aispatch marke! nuns. 6.2. Unrestrained Net Load Forecast 62.1. The unrestrained net ioad forecas! ig prepared by the Marie: Operstor by allocaling fhe foul cemancd forecast net of an initial loss percentage with respect to the Losec 4 Distr ioution Factors (LDF) set for each customer market Lf Boing AOE. 6.2.2, LDFs are defined tor gach customer marvel trading nove that belongs to a spect forecast ares, for each day and hour, 6.2.3. The iWarket Coerstor shall develop relevant procedures in the preparation of the LDFs. 6.2.4, The following figure shows the process of how unrestrained ne! load forecast: are determined from inout cemand forecesis using LDFs, fesue 2.0 -WESK-LEM Public 5 === dc2018-04-0008-pages-24.ppm === >: yes Electricity Spot Market Load Forecasting Whetmodology WESM-LFREGG Fat al diel Loon Ceccertage Saha ioitial loo Rercacrige = tk, Than Tread Loma! = 0000 ae Danni Mood 1 Custom. Wioghe 3 Saole LOS = fa Semple LO =e 6 a8 Page Aiaeetiad Lene = BOO A Jen Denepchey Led #250) ew Figure 2. Determination of Unrestrained Net Load Forecasts for each Customer Market Trading Node 6.2.5. The total unresirained ne! load foreces! ean be interpreted In the following equation: Total UNLF = Demand Forecast x (1 - initial Loss Percentage... ) Where: WAL refers to unrestrained net load forecasi Initial Loss Percentage... refers to initial loss percentage in FA FA refers to forecast area 62.6. The unrestrained net jose forecast af each sustamer market fading node Shall be obtained by pro-rating the total unrestrained ner losd forecas! to the actual joad of each customer marke! (rading node based on the latest actual systam sraoshot from the Syater Operator This is provided in the following equation: Total UNLF ee email a UNL: = Toar LDF, ea” CDF b+ Fa Where: UNLF, refers to unresinainiea ne) icad forecastof customer Market iragiwg ode Issue 2.0 AVESM-LFM Public g === dc2018-04-0008-pages-25.ppm === Wholesale Electricity Spot Market Load Foreéssling Mathocdolsgy WESM-LEM-0a 6.2.7. 6.3. 6.3.1, 6.3.2, 6.3.3. 6.4. 6.4.1. Total LOF Load, -s refers io the sum of load aisinibution factors fer hour h inca forecas! area FA where custemer market Irading node b betengs LDF, 5 Fa refers to the joad disinbution factor of customer marke! irading node o for hour) in forecast area FA A Trading Participant may submit his own woresirained nei load forecast Furtner details on customer forécasts are available in SECTION 7. FResTaaivep Nev Loap FORECAST The Restrained Net Load Forecast! of each Customer Markel Trading Node is achieved affar the optimization process of tha Market Dispatch Optimization Model (MDOM). A Customer Market Trading Node's projected load will be shed should its Drice reach the level of the Nodal VoLL orice. The Fesiraned Net Load Forecasr shall then be obtained afier the MBOM determines a Solution reflective of Inad shedding. The following equation shows the value of the Restraincd Vel Lose Foreces!, RNLF, = UNLF, — Load Shed, Where: RNLF, relars to the Aesirsined Net Load Forecast of Customer Market Trading Node b UNLF, refars to the Unresitrained Net Load Forecast of Customer Market Trading Node b Load Shed, refers to the amount of load to be shed at Customer Market Trading Node b INITIAL LOSS PERCENTAGE The Initial Loss Percentage is the percentage of the Oeanigng Forecest initially assumed to be the loss. It shall be natted out of the Demand Perecest in order to achieve the Total Linresirained Net Load Forecast, which shall then be used to obtain the Wrresiraines Vel load Forecast of each customer market trading noge issue 2.0 SW ESH-LFIM Public 14 === dc2018-04-0008-pages-26.ppm === Wholesale Electricity Spot Market Laid Forecasting Mathoadmiogy WESM-LFM-0a 6.4.2. Each Forecas! Area has its own Initial Loss Percentage. and it shall be reviewed on an annual basis by the Marter Operalor, The Initial Loss Percentage per forecast Anca shall be published-by the Marke! Operator in the Market Information Weosite’, F.1. oe] 7.1.3; Te ba, 7A. Pans SECTION 7 CUSTOMER FORECASTS a CUSTOMER FORECAST SUBMISSION Each qualified Customer, except those that are required by the Marte? Clperaior based on Section 7.1.6, may at its option, submit forecasts for its respective mancet imaging mace for each hourly or oispatch interval . A Customer may only submit-a load foreces! fora marke! tradi: 1? Fede that is solely registered to them. pict Customer that submits joad forecasts for the marke! projections and the ea-ime adisnalch shall upload the relevant forecast in accordance with the ESM timetabie, The Customer forecast snall rapreseni the estimated aggregate demand at the Market trading node and shall be in real-power quantities (MW). Customer forecasts submitted by Treading Participants shall replace the Ves ained el Load Forecast datenninad in Section 6.2 of this. Marre! Wane, Should a customer marke! trading node have a definite MW loading that can maternally affect WESIM's pricing and scheduling, then the Marke Operator shall require the relevant Trading Particioant to submit Customer forecasts for the identified market reading node for each interval in accordance with the LYE SA fimetania, Submission of Customer Forecasts in the “ESI shall be made via fhe Market Management System's Market Participant Interface (MPI). ‘hittos Awww weer, phi lesure 2 0 -WESHM-LEN Pubs i === dc2018-04-0008-pages-27.ppm === >! OS Wholesale p Electricity Spot Market Lead Forecasting Methodology WESM-LERAS ¥.2. Customer Loap Forecast ConsipeRaTiONS (ating Particpants submitting customer josc forecasts shall consider the following when determining Vel Loac Forecasts for their relevant marker traaing nodes. 1. Customer Forecast Methodology. In preparing the (osc forecas! the Customers May select the methodology that they believe is most appropiate for each Indrvidual market tracing node <. Customer Load Forecast Adjustment. Load adjustment factors shall be established by the Cusformer and his Adeter Sanwce Provider The adjustment factors shall consider site-specific losses and other factors deemed necessary to reflect accurate cuslommer load forecast al nis market trading node, 3. Customer Forecast Timeline. Customers may be able to provide their own a0 florecas! based on ihe Open Market Window defined in the WES Dispatch Protocol. SECTION 8 LOAD FORECASTING PARAMETERS 8.1. WEATHER DATA 8.1.1. Historical and forecast weather data shall be made available to the |farke: Menagemen! System's database on a reguiar interval in accordance withthe W/tSi firnetapie The weather data shall be based on the major load center located in the ‘oreces! area, Weather parameters shall include, but not limited to the following. * Temperature * Humidity 6.1.2. Such weather information shall be obtained from a reliable source deamed appropriate by the Merkel Clperafor. Prospective weather data providers are listed in Appendix E 6.1.3. The ‘arie! Ooerefoy shall maintain a repository of the historical weather- related information lgsue 2.0 VES LFM Public 12 === dc2018-04-0008-pages-28.ppm === | fp Wholesale Electricity Spot Market Load Foracasiing Methodology WESN-LFMEOG 8.2. DEMAND 8.2.1. Oemand shall be obtained from the Network Servicé Providers’ SCADA 6.2.2. Bigg. 8.2.4, 8.2.5. o.d, 6.3.1. B.d.2. 6.3.3. 6.3.4. system, and it shall be incorporated in the data provided by the Sisiem Onerator to the Warkear Ooeratar, Real-time information at the nodal levels shall also be retrieved by the Markey Jeeraior from the Sysien Operator, Such information shall be used by the Merkel Qoerater in prepanng the Demand Forecast and: tha Net load Forecast In the absence or failure to provide reliable real-time information, the (darker Qoerstor shall utilize the latest information provided, Gemand obtained over time shall form part of the historical hourly load information that shall be used by the Adarke! QGoersfor as inputs to the load forecasting algonthms used for marke! orojections and real-time cisoatch Customer Demand Forecasts shall be based on the provisions under Section Pa POWER SYSTEM CONDITION The Market Operator shall consider the forecasted power sysfem condition in thé préparation of met oad foreces!, These shall include planned maintenance (transmission, sub-trangmission, and generation), large custome mainienance schedules, and other activities, including generator testing and commissioning that may have an impact on the preparation of market ondjections and reai-hine olspatch The System Coeraior shall ensure that all relavant information ts provided to the Marke! Qoeratior in accordance with the WESM timetable For unplanned power system conditions, the System Operator shall ensure iat all relevant information is timely provided to the Mane! Managemen System through the Energy Management System or other means of communication to the ‘varke! Operator, All Trading Padicipants shall also ensure that Information regarding any unplanned activity or condition is timely provided te. the System Onmenstoar. leeue 20 VW ESA eA Pubic 13 === dc2018-04-0008-pages-29.ppm === Wholesale Electricity Spot Market Laid Forecesing Meihodalagy WESALFRAGG 6.4. Loan PROFILE 8.4.1, For similar day load forecasts, ine (Wake! Operator may utilize static profiling, proxy day profiling, and estimated or deemed profiling or a combination ihereol, The Marke! Ooeraior shall use all reasonable endeavors to provide the most appropnate load profile in determining demand and fel iosd forecast 6.4.2. The Markel Coevalor may consider typical day representations for any day of the week and season combination. Static profiling shall consider day of the week profile, holidays, and special events profile. The Marke Oneratar should note that static profiles do not reflect operating condifions of the day being estimated, 6.4.3. The ldarket Operaiormay consider selecting & day in history that most closely matches the day being estimated. The proxy day can be chosen based on aither on weather forecast and/or system load. Actual data from the sample for he selected proxy day shall then be used to create the profile. Proxy day profiling shall consider day of iné week profile, holidays and special events profile. 6.4.4. The Markel Cperator may obtaln Demand Forecasts using engineering estimates, f and when necessary, for vary predictable loads sucn as street lighting or residential area from Tracing Partiqnants 6.5. Loan Grow TH 6.5.1. The ‘danke Operator shall consider the projected growth of load demand for each Forecast Area in thé preparation of Mei Logo Forecasts, For consistency, the iWerke! Operator shall consider information from the Denatmen of Energy (DOE) for the load growth values. SECTIONS LOAD FORECAST AUDIT AND PERFORMANCE MEASURES _ | 9.1, The Ware! Opersior shall log all transactions performed for the determination of ne! joad foreces! including the uploading of the optional Customer's load demand forecast. 9.7. The /idanre? Goerator shall monitor thé actual and jorecasted nodal load for each relevant marke! trading node ssue 2.0 -VVfESA-LFM Public 14 === dc2018-04-0008-pages-30.ppm === GS Wholesale Electricity CS Who Market Load Forecasting Methacolagy WESHM-LFM-O3 9.3. The Market Operator shall evaluate the Ne! Loac Forecast Performance. 9.4. The Marke! Openstor shall evaluate the Gemengd Freres! Perormance pursuant to the performance standards that shall be maintained pursuant to VWESM Fules Clause 10.4.102. | SECTION 10 AMENDMENTS, PUBLICATION AND EFFECTIVITY | 70.1. Review and Update 10.7.1, Pursuant to WESM Aujes Glause 3.5.4.6, the Marke: Operator shall periodically review the iiss am for performing nodal igad forecasts and update this ‘arkea! Manuva! as necessary. 10.1.2. Any amendments or revision to this (Market Manuva! shall be approved in accordance with Chapter 6 of the WES// Rules and corresponding Marker MaNUa On rules change process. Vie. Publication and Effectivity The publication and effectivity of this (fsrcre* Adsnues/! shall be in accordance with Ghapter 8 of the WESM! Rules and corresponding Marke! Manual on rules change process. (eine 2.0 —W¥ ESM: LPM Public 15 === dc2018-04-0008-pages-31.ppm === «>: CES Wholesale le Electricity Spot Market Load Forécashng Mahocnogy WES H-LFB-O SECTION 11 APPENDIX Appendix 4, Weather Adaptive Algorithm for Short-Term Load Forecast 1. lppuT Data SPECIFICATION The following input data aré necessary io calculate the (osc forecas/ for this methodology: a. Current load data b. Gurrant weather data. 6. Weatner foretast data provided by weather senices companies. d. Preselected day types for foracasting period. @. The prediction error of the tast forecast. i. Estimated parameter values which quantify the causal relationship among nistoncal load data, historical weather data and day type data. 2 WEATHER ADAPTIVE USING MULTIPLE REGRESSION METHOD The Weather Adaptive Algoritnm uses multiple regression analysis to create a statistical relationship between demand and weather conditions, including day types. In this methodology, the regression coefficients aré estimated using equally or exponentially weighted minimum mean square error (MMSE) estimation criteria, The regression analysis is performed with up to 3rd degree polynomial functions. The polynomial degree is adjusted for each weather influencing variable by parameter. Regression analysis fitters the different weather-dependant parts of the load for a given time, ior example, time mierval 4.00-9.00 a.m., as well as the influence of day type. The result of this filtering process is a set of regression coefficients (polynomial-coeftficients). To detect non-typical historical data, an adjustable data check is implemented, i load or influencing variables violate an adjustable bandwidth, they are excluded fram the data analysis, The result of the analysis is a decomposition of the time-dependent total load Yi) into: issue 2.0 —WV ESL Putodic 16 === dc2018-04-0008-pages-32.ppm === C2) Moctesale! Electricity Spot Market Losd Ponecasting Mathedalsgy WESM-L FoF F(t) = Alt) + apT(t) + alte) + [> ayW'(t)| + Of) + model_errar(e) Where: Bit) refers to weather insensitive load component a, L(t) reters to weather-sensitive load components ap, T(t) refers to temperature-dependant load a, Lit refers to humidity-cependent load ay Wit) refers io other weather-dependent load Dt) refers to day type-specific load component model_error (t) refers to the minimized weighted least squares estimation The model error, is minimized using a weighted least squares technique. and this allows ine user to gel the forecast results in the form of an analytical expression. Generally the regression analysis, which involves regression coefficient estimation, ig activated periodically together with a subsequent forecast calculation. a. MODEL ANALYSIS Model analysis is the process used to find a néw set of parameters that reduces the forecast errors for the weather adaptive algorithm, This requires using a panicular set of nistoncal load and weather history. Multiple simulations aré tun wich imitate what the Weather Adaptive forecasting calculations would have been for that particular sel of data starting at @ user defined day and moving through the ramaining data one day ata time. Each simulation has a different set of weather adaptive parameters, Results aré summarized and may be compared, and therefore help to in analysing points that need further improvement. 4, DETERMINATION OF THE WEATHER FORECAST DATA In order to gat good forecast results the best available weather forecast data are NeCeSsary. For the forecast calculation, the STLF program requires time series in the adjusted time step of each weather vanable over tne whole forecast period. These values are calculated automatically from the preceding inputs thal are listed. issue 2.0 -WVESM-LFM Pubes if === dc2018-04-0008-pages-33.ppm === Wholesale Electricity Spot Market Load Foncasting Meatadaiagy WESML PREG For ihe temperature influence, the program tracks a standardized curve through the day-specilic values of minimum and maximum temperature. This standardized temperature curve is determined from the historical data at each run of the analysis program. The time points of the extreme forecast température are extracted from the historical data. 5. DETERMINATION OF THE LOAD FORECAST The joad forecast is calculated for each adjusted time step of the forecast period using the regression coefficients and the weather vanables forecast as wall as type ot cay, The forecast program is activated through the following different means » Pernodically al defined time steps » After each run of the analysis program » On user request Igcue 20 4VESM-LEM Pyblic #5 === dc2018-04-0008-pages-34.ppm === Wholesale Electricity Spot Market Load Forecasting Methodology WESH-LFM-OS Appendix B. Similar Day Algorithm for Short-Term Load Forecast Use the similar Gay forecast to produce a Oemand Forecast fora day or a range of days in (he forecasi period The similar day forecast is based on recalculated Oemarct patterns. There are Demand increment patterns for each - Forecast Area Demand - WMenth - Hour minute - Day type - WNormal/average weather conditions The similar day profiles where calculated initially using historical Demand data and recalculation is going to be done once a day during midnight processing. Subsequently, they can be reviewed and edited using the STLF User Interface. Historical load profile (similar day load) has been calculated for each month, hour and day type: 1 " LoadProfile, ij. = in > HistoricalLoad, ,y f= And standard deviation for each caimuvated load profile has been calculated by: Standordieriation y= penitence fea Where: pat. 2. 3.12 = month \=1, 2, 3,,..24 - hour K=1,2,3,...8 - day type [=1, 2, 3....0 - number of historical days Forecasting applications supporting the following day types: (Day Type No.) Day Type a | 2 MON 3 SAT | 4 SUN 5 | WKD less 2.0 -VWESH-LFEM Public 718 === dc2018-04-0008-pages-35.ppm === &>: ehekenle e Bleckricity Spot Market Load Forcasting Melhodalogy WESM-LERMAIS Day Type No, Day Type | | 6 THE 7 WED 8 | THU The Cemand similar day patterns are non-weather sensitive, and they provide the average Oemand in case for normal weather condition for that time of year (defined by month, hour and day typei. The weather increment pattems defined by standard deviation are expected to change under “Hot'/"Cold" weather conditions. Positive increments indicate that load increases in hotter weather conditions than in normal weather conditions, Negative increments indicate the reverse. Under cold weather conditions, the incraments are subtracted rather than added to the non-weather sensitive load pattern values, leaue 2.0 -\WESR-LFEM Pubic au === dc2018-04-0008-pages-36.ppm === > wlesal Electricity Spot Market Load Forecasting Methodalngy WESM-LFM3a Appendix C. Pattern Matching Algorithm for Short-Term Load Forecast The pattem matching algorithm uses a process to filter days that can be used to determine Demand Forecasts. The filtering of days is based on a@ pre-defined set of parameters, namaly the type of day and historical weather data. Based on tne values defined in the filter, such as the type of day and the high and OW values of temperature, it will select days and then provide an index for tts differance. From there, it will assign a weight to be applied to the Demand for that day. Using the weights along with the actual Oemand tor that day, a Demand Forecast shall be determined. Issue 2.0 “WESM-LFM === dc2018-04-0008-pages-37.ppm === &>! Wholesale Electricity Spot Market Load Foracasing Methodology WEST-L FAH 4ppendix D. Cubic-Spline Interpolation Algorithm for Very Short-Term Load Forecast 1, Cugic-Se.ine INTERPOLATION Process The §-minute Demand Foreceas! axecutes for a time-honzon of 2 hours, or for 24 5- minutes intervals. The §-minute forecast-area load forecast makes use of a Cubic Spline Fitting algonthm in conjunction with the load-tracking methodology. The inputs to the calculation are the hourly Oemangd Forecast as well as the 5-minute historical loads as shown in the figure below, Cube Spans Fig + Tacrg Sree Demard Figure 3. Gubic Soline Process Overview The Tracking algorithm ¢altculates correction term ta the forecast for several 5- minute iorecasting period. A curve fitting algorithm is used along with acival Oemeanc to calculate the 5 minute forecast. 2. 5-MINUTE TRACKING LOAD FORECAST FUNCTION Wid CUBIC SPLINE The inca. forecast if based on deviations between the five minute Domanc E esis and tha actual Deciond every five minutes over the last hour, An average error -tarin is used for the tracking forecast calculation. This average error term is calculated by adapting the relative errors for each time step so that values reflect the recent érror trend, _ |SMinuteLosd, — SMinuteLoad, | f= SMinuteLoea, [soue 2 Oo —WESA-LPM Pugalic Ze === dc2018-04-0008-pages-38.ppm === fp Wholesale Electricity Spot Market Load Forecasting Methodetosy WEEM-LFM-03 A =A, + ae, —A,, | Where: Siinubeload, - § minulé lorécast load at time t SA Iniitel oad «6 minute measured load at time t é - Aiglative error at time t A - Avetage (or adapted) error at time t a, - Adaptive parameter (0 1) The adaptive parameier ls selected so that the error from the previous time step does not receive too much weight. The weighting factor is introduced so that the efiect of ina error jor the previous time step may receve additional weight in determining the correction tor the next time step. issue 2.0 -VESMLFM Purhiic Pa === dc2018-04-0008-pages-39.ppm === Ges) Wholesale Electricity Ge) ho Market Lita Ferenaciing Meinacdoalocgy WESM-LFMLOS Appendix E. List of Prospective Weather Data Providers 1, Depanment of Sciance and Technology — Philippine Atmospheric, Geophysical, and Astronomical Services Administration (DOST-PAGASA) 2. Accuweather lesue 2.0 -4'ESM-LFR Pilla 34