Loading...
HomeMy WebLinkAbout0167.094, . ~. •.•. . .:~ - February 17, 1994 Honorable City Council Polo Alto, California MmDben of !he Council: R=ort iD Brief . ,,",,---.~',: .~. This report. approved by !be Ublities Advisory Commission, forwards !O Council information regarUing!he January 1994 Gas Inregraled Rcsou= Plan (IRP). No action is required at this time. Included in !he report are: 1. Information and assumptions about gas supply and transportation resoun:es. 2. A deso..-,;ptiOIJ of the analysis u'W:d to evaluate the possible ga.o:; rescNlrc.e p1alL'>. 3. Short-and Icng-rerm ",eool!nendatioDS. RprlnTvmA Integra!ed resource planning is a research activiry that resui[s in .a busmes.... action pian. Information is ga!hered and a:JaIyzed !hrongb a !II1K:lUred process 10 find !he relative costs 01 many plans. Palo Alto', Gas IRP is an analysis of plans involving both supply and demand-si(k resources. The goal of this IRP is to develop a framework for decision making in !be gas utility. Tho plaJming process was inmate<! to determine wbelher me City should make " IS-year commitment to !be Mojavo pipeline, • propo<ed intrastate gas pipeline that will bypass Pacific Gas ok Electric. To address !he problem, staff evaluated all of !he gas supply and transportation alternatives available to !he City. The poIOntiaJ gas resources included: Southwest US andJor ~ supply via five different transpOrtation options, demand-side management, and CMR:167:94 • i "'-.. ,--/' ~~' - ownership of CaliforrWt g&S production. Two cliteria were used to cJlta..."llirn: each pl:m. to[a] resource cost and rate impact Utilities Adyjsoo: CgmmjWpD Action The January 1994 Gas iRP was presented to the VAC OD J.ruary 5. 1994. In the preceding year. tile UAC has examined tbe Gas iRP as it has progressed fro", !be estabhshmeru of ~ alternatives and evaluation criteria to its current: form. At the December 6. 1993 meeting. the UAC made a moti(m £0 accept the Gas Integrated Resource Plan and 8.C'..cept within it tbe short-lenn action plan on page 27. It was found that Southwest US gas via the Mojave Pipeline and one other ajlemati¥e Canadian gas via thc P-""ific Gas Transmission Pipeline and PG&E system, are equally attractive. For thaI reason S!alf will continue the negotiations wilh both Mojave and PG&E. Staff t.'1en '" ill issue a report with a fmal recommendation at the conclusion of the negotiation process. Re.pectfuIly submitted, C-~+\-..1,,~ TOM HABASH! Manager Resource Planning l{n.·j EDWARD;~ Director of Ulilities ~~.~ BERNARD M. STROJNY Assistant City Manager A_: January 1994 Gas lruegraled Resource Plan cc: Utilities Advisory Commission CMR:167:94 / lilY OF PAlO ALlO G4S UIIUIY lmuary 1994 1NIEGRA1ID RESOURCE PIAN Prepared by the Utilities ])epa! tment Resource Planning Staff PrinJed 011 lIecycied P.:per Table or Contents Table of Contents ........................ , ........ , ..... i ListofFigu= ............ , ................ , ..... , ......... , iii List of Tables .. , ......... ,., ... '................. iii l..isl of Abbn:viations . . . . . . . . . . . . . . . . . . + • • _ • • • • • • • • • • • • • • • • • • •• iv Execulive SIlnlI!IDY . . . . . , . ' . . . . . . , . . . , . . ' . . . . . . . . . . . . . . . . . , . . . . 1 Cbaptcr 1 Background & Context of :be IRP . , . . . . . . . . . . , . . . . . . . . . . . . . 3 1.1 Historical Background .. . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . , . . . . . 3 1.2 Motivation", do :be Plan ....... , ' ....... , .................. 3 1.3 _ Surrounding Gas PWming .,.,., .................... , .. ' . 4 1.3.1 FERC VS CPUC R£gulation . . ' . . . . ' . . . . ' . . . . . . . . . . . . . . . . . 4 1.3.2 Copadry IJroko'inglReleau .....................,........ 4 1.3.3 Oth<r Proaedlng, . , ........ , .. ' .... ' .. , .. , ........... 5 1.3.4 Cal/Qila/CaIi[omill Dispwe ............................... 5 1.3.5 Moj~ Pipeline ..•................ , ........•........ 6 1.3.t; Gas Dmumd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.3.7 Noneo,.. Quromen ...... , ........................•... 7 Chapter 2 Cbarocreristics to Appraise ............................ .... 9 2.1 0lU ............ , ....................................• 2.2 Relillbility ........ , ....... , ......... , .................. 9 2.3 Gas Supply Options ....•...... , ........................... 9 2.4 Storage. . . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.5 Operational CoDCeTll> ...................................... 9 2.6 RcguIatory Sensitivity ..•.... , ... , ............. , .. , ........ 10 2.7 Availability ........................................... 10 2.8 Experience ............... , ........... , .. , .. ' .......... 10 2.9 Administtation .................. , ...................... 10 2.10 EnviromnemaI Impacis .....,.............................. 10 2.11 Robustness ................ , ............ , .............. 10 .::-', Chapter 3 Qualitative Assessmeru of Alttrrunives . . . . . . . . II lnb:rstau: Gas Supply and TraIL'!""",,;oD ......................... II PG&E Allemati ... es .................................. 12 100% PGf ................................... 12 50% PGT and SO% EI Paso ........................ l3 101J'JI, EI Paso ................................. 14 CQ" Subscripri<>n . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Moj"'" PipeIiM .................................... 14 3.1 3.1.1 '.I.U 3.1.1.2 3.U.3 3.1.1.4 31.2 3.2 Ownmhip of Califomi.1 Production. . . . . . . . . . . . . . . . . . . . . . . . . . . . . IS 3.3 Demand Side Mcasu= .................................... 16 Cbapter 4 Decision Analy,is .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1 Problem Formulation ..................................... 17 4.1.1 .Ailematives •. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.1.2 Un.."mi1inti.. . ..................................... 19 4.1.2.1 Gas Demand .................................. 19 4.1.2.2 FinanciIJI Rat .................................. 19 4.1.2.3 Gas Prices ...............•......•............ 19 4.1.2.4 Pipe/iN! Commilmenl Level and Rat.. .................. 19 4.1.2.5 Slorage .................••.................. 19 4.1.2.6 Orho Mcjave Costs ............................. 20 4.1.2.7 DSM •...................................... 20 4.1.2.8 Drilling ...•................................. 21 4.1.2.9 R£guIa;ory and Administrative ....................... 21 4.2 Dctcrministi<: Analysis .........•..............•............ 2. 4.2.1 l'Ia)Jou .............................•........... 23 4.2.2 SeruitM!)· AMlysiJ ..•................•.............. 23 4.2.3 Dezl!T'1'Jnisric Outcome and Resldrs . . . . . . . . . . . . . . . . . . . . . .. . 25 4.3 Probabilistic Analysis ..................................... 26 4.3.1 Purpose .....•.....................•............• 26 4.3.2 Expm ItllerMws and Key Variables . . . . . . . • . . . . . . . . . . . . . • . . 25 4.3.3 kruIts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 ..................................... 29 Appendix A Description of Uocertain Variable, Appendix B Details from Detenninistk S!lldy Appendix C Deuils from Probabilistic StlJdy ii --. - I • Figure I Figure 2 Figure 3 Figure 4 Figure S LISf OF FIGURES Gas Use vs Purchase ........... . SclJematlc of Pipeline System . . . . . Decision Tree, Detmnirustic p~ . . Influence Diagr.m .......... . Decision Tree, Prob3bilistic Phase ...... . .................. 7 ........ 11 . . . . . . . ... 18 ..... " . 22 . . . . . . . . . . . .... 2S Figure A-I ......................... Palo AllO Historical Gas U"'ge Figure B-1 ............. . ... Variable Sell.5itivity Tornade Diagram. TRC Figure B-2 ....................... Variable Sell.5itivily Tornado Diagram. SAR Figure B-3 through 809 ... _ . . . . . . . . . . . . . . . . ... Rainbow Diagrams Figure 808 . _ . . . . . . . . . . . . . . . . . . . . . . . . . . ....... Nominal Plan TRC Figure 809 ................... _ . . . . . . . . . ... Nominal Plan SAR Figure C-I ....................... Probabiliry Distnbution Canadian Gas Price Figure C-2 . _ . . . . . . . . . . . . . . Probability DislIfoution Southwest US Gas Price Figure C-3 ................... _. Probability Distnbution EI Paso Pipeline Rate Figure C4 .,. _ ........................ _ ......... _ Ri3k Assessmeru Figuno C-5 . _ .. _ ..... _ .................. Annual Cost by Caregory Mojave Figuno C-6 ............................... Annual Cast by Category PGT Figuno C-7 ... _ . . . . . . . . . . . . . . . . . . . .• "ulmJaI Cost Comparison Mojave vs PGT LIST OF TABLES Table 1 Summary of Alternatives ...... _ . . .......... 18 Table 2 Cbanses in Nominal Policy -TRC ....... . . . . . . . . . . . . . . . . . . . . . 24 Table 3 Changes in Nominal Policy -Rate Impoct . _ .... _ .. _ ...... _ ...... 24 Table 4 Negotiable Variables, Probabilistic Phase ................ _ ....... 28 Table A-I _ . _ ....... _ ......... _ ..... Uocertain Va,iables Deterministic Phase iii APGA CERA Co, CPUC DSM DTH PERC IRP LDC MCF MM8TlT NPV PG&E PGT ron: SAR mc WACOG LIST OF ABBREVIA TrONS American Public Gas Association California PubLic Utilities Commission Demm:! Side Managemem Deca<benn, one million British thermal units Federal Energy Regulatory CommissioD ~ Di£tributioo Company One thousand cubic feci am srandard tempetalUre and pressure One million British thermal units Net Presem Value Pacific Gas and Electric Company Paciflc Gas Transmis<ion pijleline Pacific Gas transmission expansion pijleline SySll:m Avenge Rate Total """""" cO<! Weighted Average Cost of Gas iv ------- ~ -. --_._----"------ Executive SWill my ~ resoon:e planning is a research activity that results in a business action plan. Infilrmatloo is ga!hemI and .nal,zed th.'1lIlgh a structured process to find the relative costs of lIIany plans. Palo Alto's gas lRP i5 an analysis of plans invotving both StIpp!)-and demand-side resour<:es. The goal of this lRP is to de\-elop a framework for de";si"" making in the gas utility. The ~ gas resource planning process was initiated to de!ennine "bether the City should make • IS-year eotr.mitmem to the Mojave pipeline, a proposed L"ttastate gas pipeline. Another decisioo had to be made fIrst to ansv.-er the qu<:stion, "Should Palo Mo acx:q:t a pceammged deal fa: interstate pipeline capacity from Pacific Gas & Electric Company (PG&E)?,' The la!ter dccisioo was made with the original dedsioo still imrnin<nt This tqJOrt documenIs the City of Palo Alto integrated gas = planning acti;ity that was dooe to address these iss..,.. Included in the report are: I. Information and asswnptions about gas supply and b3ilSfXJ!tB1ion resources. 2 A desaiptioo of the analysis used 10 evaluate the JlOSSlble gas resource plans. 3. Short-and long-tenn reco!IlI!le!lda. To """"'" the qw;tion regarding Mojave. dedsion analysis ~hniques were used in the planning ]2OCeSS because these methods provide a systematic ".r 10 examine complex silUalioos. The ]2OCeSS has five ""!ior steps: problem formation, model de\-e!opmen~ risk analysis, plan adoption, and action. The p!apooe of the problem furmuJaticn ~ is 10 define !he problem, gen=te a list of a11<matives and uncertain variables, and eslabiish the criteria used to evaluate the alternatives. To address the problem of ddmnining if Palo Alto should make a 15-year commitmertt 10 Mojave pipeIine, SIaff evaluated all of the gas supply and transpclt3rire aItematives available to the City. The JlOI=l!iaJ gas = Included: SooIhwest and/or Canadian SUFriy via five difIi:tent transpottation options, dcmaIld-side management, and ownership of California gas productioo. Two criteria were used to examine each plan, tot.! resou.= cost (mC) and Iate impact The analysis J'I'OC"SS began with the developmtnt of. value model ,.hic:h links the un=taintics to III ouIl:mJe so that the alternatives can be wwpaied The model analyzes all the gas resource optioos avaiIable COIlSldering the !IllII1)' UIICer!Binties that exist about the future. These UIJa2'taimies are b3IldIed by developing a reasonable r3Ilge of values fa: each variable that will affi!ct the cost of any I":SO\II1:e. The objective of the deterministic phase is to reduce the numbet of posstble plans and the rrumber of variables to only the most ,iable plans and sensitive 'Wbles. The det<rministic jiJase analysis n=Iterl in postponing the dedsions aboot demand-side managcmcnt (DSM) and -", ."~" California ptXIuaion until the nex! IRP. The OI'Jy remaining allematlv,," were the out-of-stale gas and n.nsporilltion options. This decision Wa'l found to be sensitive to thm: variables: the price of gas in the SotJIhwest US, the dis<:cunt for Canadian gz. and the market price roc EI Paso Na!waI (J&<; pipeline ~. 1b:se !hn:e variables "ere analyzed further in the next phase, the probabilistic stage. For each VlIriabIe, an expert was iderttilied to assess the iikelihood of the future v.eJue of!he variable. The export's judgomm! "as cIrIMn cut by using. fannal. struclll!td iJl!eview process. The result was a probability dL<zr:ibulion of the JX>SSible values far each variable. The model was lh6J updared U'Iing !be infonnaIioo from the expert intttviews. Once sIaff deo.-eIoped an unde:standing about the future througt! the above analysis, !be da.-isions were revisited The first decision that had to be made was v.1lether io aco:pt a ~ dell for inretS!ale capacity from PG&E Based 00 the analysis, staff r=mmended not to accept the deal. The action eliminated one of the five alternatives. WIIh !be ~ deal rejeded, Palo Alto still had to make a short-= decision, "Should Palo Abo bid for short-tmn Pacilic aa. Transmissim (POT) capacity 00 the ~ mari<et?' Staff 1"C<XlIIllIIfnd to bid for capacity rights dJrough SUITtlIIeI" of 1995 in orda-to secure pipeline capacity without eflecting !be Mojave decision. A bid ,,"" made. and Palo Alto was SIICO:<Sful in olXaining 1his capacity. The original questi<ln, "Should I'liIo Alto make a IS-year t'"tntliCild~ to !be Mojave pipeline?', still had to be answered. It "'MIS found that Mojave and one other ahemative, Ca!mdian gas via !he POT and PG&£ pipeline's, are equoUy attrBCtive. The other three a1ldrotives had net presem values approximardy 515,000,000 bigherthan the Mojave and P<.J' options. Staff !e<'OIlllDdIds to engage in ~ons with Mojave and PG&E to hRak the tie. Some quamitative infamatioo was inlcmiooalJy omittai from this r<pOIt to preserve the oonfiderrtiaJ~ of !be negotiaIioos. Staff will issue a report with a final iCCOOJIfidldation at the cmclusioo of the negotia!ioc process. 2 1. BacI!gmWld and Context of the IRP 1.1 Hstorical BIrlgrolind Since 1986, the Federal Energy Regulatory Commissioo (FERC) and State Commissions have been wcOOng toward the dtregu!atioo of the natural gas busin<ss A new FERC rule is in place whid! ",ill complete this deregulation by making interstate pipeline capacity, whid! has been monopolized by big local distnootion companies (LOCs) like PG&E, available to customers like Palo Alto. Until 1991, Palo Alto relied aIrnos: entirely 00 PG&E for its gas procurement and tzallsIX"tation needs. t, 1991, an interim program was put in place by the California Public Utilities Commission (CPUC) to allow "holesale and industrial custOmeTs to procure gas supplies on their O\\n and use the !.lXs' L= pipeline capacity to IrlInspO!t the gas to the California border. Palo Alto successfully participated in the IXOgram by ~ing Jow-cost Canadian gas from Shell Caroda and using PG&Ps pipeline riglus to lIallsport it In AllgIlst of this year Palo Airo signe<l a new gas supply ~ wi1h another Canadian produ=, Suncor Inc. Suncor holds pipeline cap"dty 00 the Pacific Gos Transmissioo e:.pansioo pipeline (PGTX) mud! is • new line from Canada to the California border. Suncor ~ the gas to CaJilixnia. Palo Alto has inttastate tt31lSPOltation service from PG&E to br'.ng the gas to the city gate. 1.2 MI1ivaIion to 00 the Han For the future, the City has some inl=sting opponunities. AIthougb Palo Alto always bas the choice to become • PG&E customer again, the City has the ~ty to acquire intasIate pipeline capacity of its 0'Ml. FERC Order 636, along with the CPUC, capacity broIrering program, provides the framework for inu:rstate pipeline capacity to be released by PG&E and SoCaIGas to other parties. Capacity on PGT and El Paso Narun!I Gas (El Paso) Pipelines is ",'lIi]able through the capacity brokering JrOCCSS. In addition, the MJjave Pipeline Company has propos<xI to build an extension of their pipeline from the Arizma border into Northern California, completely bypassing PG&Es system. TIlIIing is critkal as Palo Alto will need to make a oommitmern to MJjave by early 1994. 3 ,. 1.3 lssues Swromxing Gas Planning 1.3.1 FERC Vemus CPUC RegWIIIon EVal thou!#1 Palo Alto's gas utility :s regulated by Cit:o' Couocil, the ",,"!ions of the CPUC and the FERC affect !he gas utility • great deal, To insure that Palo Alto's interests are not overlooked, Palo Alto has participated in all ",-levant proceedings aI !he CPl:C and has become more involved recently with FERC issues. These a::tivities take a great deal of staff and consulting time and IDOIII')'. In addition to Resou= Plaming SIaIf spending time on these issues. the City AtIoole)Is ofike also carries a large burden to stay abreast of the happenings in the regWaIay arena. All gas transportatioo alternatives before us require the util ization of at least """ interstate pipeline. That ClCIIlS Palo Alto will be con=ned ,.ill! FERC decisiom regaxdIess of the transponation al1<:rnative chosen. To dale. Palo Alto has not spent nu:h time appearing and testifYing aI these proceedings. Staffbelioves this srtu.tion will remain SlalUs quo becalIse (I) The FERC moves sJowl)' and does not make &equern ~ to !he reguhII<xy structure, (1) Oz interests are indi=tIy JnlIeCIed be::ause of the huge IUIIIlber of municipalities IICro!S the COllIllIy that participllle at !he FERC, (]) Palo Alto already participates through membership in the American Public Gas Associarion. and (4) It "ucld be • wa<le of time and money Ie participate as a stand-alOlle party in any procoeding althe FERC unless there is a very oanmv fo."" on NorIhem California The CPUC proceedings affect Palo Alto through those alternatives that rely on PG&E's inIrastate grid fur trnnspottation to !he city gate. Unlike the FERC, the Califurnia Commissioo reguJariy holds proceedings to rev;.,. and change raIeS on the PG&E system. Palo Alto is unique in that it is the unly municipalutilily on PG&E's system that participau:s in the prooeedings, A great deal of time and money are required Ie aae:I4't Ie protect Palo Alto's interest at the CPUC. In fact, nearly our entire consulting bu:lg<t of $] 50,000 (approximately 6% of the total it-. llimsponalion oost) is spent in this fushion. UnfurtunaIely, Commissioo decisions are most oficn unfavorable to Palo Alto, and in many ,mtances the City's conccms have been ignoml all togelher. eap.;ity~ Regulatory issues are !he most complica!ed and unc<rtain aspect of plonning for Palo Alto's gas utility. As discussed in !he intnxho:tiun, the regulatory framework of the gas industty is in !he midst of. tadical change. Mmy uncertainties remain aOOut how the marlcet will behave in this new stIu<:ture, The FERC has issued Order 636 which lays out guidelines for interstate pipel ines 4 to release capacity on an "J."!l marl<et. Interested CUSIomer.'i rna) bid up to the maximum approved rnte for space en e:h.i1 ~cuIar pipeline. PG&E is releasing capacity on PGT trom Canada and on El Paso from the Southwest U.S. However. Onler 636 00es allow LDCs to make pre-arranged deaIs to release capocity before it goes to the open market In this ""'" the CPUC rules come into play be<:all'le PG&E and SoCalGas are the main interstale capocity holders in California, and the CPUC regulates boIh of these entities. Palo Alto can reserve capacity for i1s core (resid<ntiaI) load through a pre:nnmged deal wi1h PG&E. H:lwever, under this option Palo Alto is required to split i1s capacity equally between the PGT and El Paso pipelines and pay the full as-billed rate on boIh pipelines. This could make good gas supply corura:ts harder to negotiate since our load is already moderately su.d and extremely seasonal. 1.3.3 0Cber I'rtla!e!tDgs The CPUC has sever:;] other proceedings un,""",,=, ,,11iob Palo Alto has been evaluating and will continue to monitor. The fu.t one concerns the unbundling of PG&Es stornge senrice. This "ill result in tMklng sL.-.-age avaIlable as a separate serviee from the intrlIstare transponation servke. S~.lI unhundl ing may l""'ide customer> with the ability to store gas on SoCalGas or Ihird party storage S:JStems. CUm:ntly, !'rJ&Es tranSpOrtlIIion :ate includes stcrage cc&s. Another proceeding which has just been announced "ill make inlrnstale capacity .... 1IiIabIe through a lxukering mechanism. The plOceedillg has not begun, so no infurmaliOilisavailabIe. The Biennial Cost A1100ation Proceecfmg (BC."J') determines bow lIlll<:h of PG&Es revenue requirernertt will be aI1oc:ned to each customer. The next BeAP will be in J 994. The Long Run Marginal Cos! (LRJvIC) proceeding set guideli:1CS foc a new m<!hod of allocating cc&s to FG&E custorncn. Rates are oow based on the Ioog nm matginal cost each ctJSICmer causes !'rAE to incur. U,4 ClmadaICaIifotD Di5pJIe For SC\'e!3l years, there has been a dispute between Omada and California regarding price and availability of gas supplies. This is an extremely complicated issue. Canadian gas has been sold under cootract io CaIlfani. by a distinct group of producers knov.n as the AIberIa and Southern Pool (MS). The price of gas to California has been based on a "neIback" pri<:e. This means that the price of gas has been calculated by subttacting the cost of trar>;portation 5 from Canada '" California from the price of U.S. gas at the California border. The result is !hat the cost 10 California customer> is ..tmost !he same v.he!h<r the !!BS comes from Canada 0< the Soothwest u.s. "-'""'PI for • small discourJ by the Canadians. Many have described the si!Ualion as a carteI because the A&S gas that ends up in California is sold at a substantially higher price than the gas tim goes to the local Canadian market. Wbtn the CPUC issued its capacity brokering decision, the Commission de<:lared these CC!1I!ac!s to be null and "oid and argued that steps !l1lJSt be taken to J1X)\"< toward. competitiyo marke! = quickly than orisinally planned. Naturally. the Canadians have not re5JX1!1<Ied liIvorably to the CPUC's disregard for these cootta:ts. Negotiations have recently concluded v.hich ddmnincd !!ow nwch PG&E will pay 10 n:st!UC!Ure the Canadian COtttr...:ls. This cost was "W'u"imately S250.000,OOO and "ill be passed on the rate poym and stock holder.; of PG&E and the PGT customers. MljIwe PIpeIino As memi<n:d, Mojave is a jlI\lIXl5ed pipeline from the Ari20na border 10 Northern Califumia. The success or faIlure of the project bas enonnoos implicat..'ons to the CaIifornia gas mada:t. Mojave "ill be an int<rstate pipeline regullIIed orJy by the FERC. Eve!)' =aomer that switches from PG&E to the Mojave system will cause a greater 1uden (II !boo-remaining (II PG&E's system In order to discoInge bypass (alSlomo:rl; leaving PG&E and the CPUC jurisdiction,) the CPUC has approved sevemI discounted contracts as an incentive to stay. The problem is that the CPUC can not bind furun: Commissims to honor discoumed rates, so there is 110 gwrB!IIee !hat the discoun!s won't disappear in the future. Palo Alto is in the same position as all porentiaJ bypass CUSIo!'!ler.i. Two good 0UfC0ttIQ; "'" possible: (I) Mojave is buil~ the project is cost<ffectiYO fur the City, and Palo Alto is coonected to it; or (2) Palo Alto =iYes a <fuw.Jrued rnIe from PG&E as an mantive '" SIa)' on the system Anocher possible outcome is that MljaYO is built, Palo Alto does not comect to Mljavc, PG&E does not offer a discomted rate to Palo Alto, and the City ends up paying stranded costs caused by the b).poss of other PC'..&E customers. Palo Alto docs believe, however, !hat the pipeline 00IIIfCIiti00 that Mojave "ill foster will benefit the stale of California and every CIl:!toIIII:r (II the PG&E system. Thc:refure, the """l! po!lSlble outcome is the iiIilure of the Mojave proj~ and Palo Alto has participated at the stale and fode:allevel in suppon of the pipeline expmsion. l.3.6 Gas Demoad Palo Alto's av..-age gas demand is approximately 10 MM:FD (MilIioo cubic em per day.) This varies si!Jnificantly beIv.een summer and winter with a typicaI 6 , . average July usage of 5,000 MCre and .... crag<: January usage of 19,000 MCFD Annually, the purchase profile could malch the demand profile depending on the amDUil! of storage available 10 PaiD Alto, Figure 1 sho\\~ Palo Alto's gao purcl!ase profile ",til and "ithou! S!O<age for 199:-1994. City of Palo Alto N.Jn.jrai G.u Use .a!1d Purchase Profile ~".,.. 18 , 16 , 14 t '" ,. : .. 10-~:~. 4 ' 2 ' 0 -----_ .. _- f e. 1S > " .. 0 ~ • ., c i -;; .: -, .. ~ • <: • W Flg1n I Since Palo Alto is 110! c::<pcriencing much grov.lh. healing degree days are tlte most iJIl!XlIcmlI variable in detomining the City'S gas demand. An econometric modcl using regression as the forecasting method "'th bigh and low boundaries set by two standa:tI de ... iatiOllS was used 10 de-. ~lop the demand f=cast. Howeve:-, the IRP docs not tten demand as • fo=ted ,alue insteaC, it is treated as an 1II1CeI1aimy. Palo Alto's historica! gas usage and !he r:mges ... hlch """" resred in !he decision anaJ:.~is process are shown in Figure 4, in section 4.2.4.1 1-3.7 Nooro~ 0Alme0s ApproximaIeIy ten pe=ru of the City's load is <:OO1prised of YlOIlCOO! =. These custom:rs receive • dis<:ourned rare bo:tause of their sta!Us as ar.ailable custome:s "'th aitema1C fuel capability. Staffhas llIlderuli<en discussion "1t1l the noncore customers to make them aware of the !le\\ regulato<y environment. Some of these ~ have expressed that they "1lIII to cootinue to be served 7 ". '-' by Pak> Alto Mille others have expressed the desire 10 acquire pipeline capacity and/or gas supplies on the'.r cr.' n. Resource Planning and Rates have been working IOgdhcr to establish a ~' 10 provide ttansportation-only service 10 lOose customers who d<:sin: it. 8 2. ClJar.k:teristics 10 Awrnise In order to evaluate each altematiVe. 3 I ist of eleven criteria wen: developed. Each one is dc:la1bed below althougll most are .. If explanatory. Each of these <harncteristics were coosiden:d during the development of the value model in the problem fomrulation phase of the dedsion analysis process. This list was the basis for deter:minir.g the costs to be induded in the = cost of each alternative and the uncertain variables to be tested in the dctetministic phase. 2.1 0JSt The cost< (both present and future) associated wiTh eacl: alternative include: (1) the inrcrstate and inttastate tran.sportalion charges, (2) gas commodity costs, (3) investment or capital costs, (4) opera1lng <0Sl<;. (5) storage, (6) back-up, (7) other ='1arges, (8) regulatory and litigation costs, and (9) adrni.-rlstratP.'e costs. Unit costs :m: usually expressed in $ l"" decathe!m. 2.2 Reliability Reliability is critical to the City. The reliability of a pipeline system may be assessed by its configuratioo. the number of lines serving the area, etc. Back-up service ill case of an emagency may need to be as.sc:ssed for some altematives. In addition, reliability provided by <XlIIIraCt = 'With the pipeline. gas supplier. or oIher end-users is possible, 2.3 Gas SURiY q,ti1llJl The sou= of gas supplied via the variety of basins, producers and marXeters and made available tbtou£h any of the transport&ion alterr.atives is important. The two main regions providing gas supply to California are Canada "h.idt includes gas from Alberta and British Columbia and domestic supplies including gas from the Permian basin, the San Juan basin, and the Rocky !l.-Iountains. Califomi.a gas also contributes some supply to the state. 2.4 SIDnIge Storage can be considered '" reliability insuroix:e if it is used to supply gas to Palo Alto at a time of cunailmcm 011 the int=td.e pipeline. However, Palo Alto also u!Jli= storase to fIattcl the Iood rrofile and to buy inocpensive gas in the summer fur winttt use. The storage o¢oos =ociln<d with eacl: n' .. ,pouation aItanativc rnu<;I be coosid=red. 15 ~ofiooaI o.nr-.s Any change to the operatior. of Palo Alto's gas distnbutioo S)SIem is criticallD be aware of and to include in the evaluation. Such changes may include • diff=n number or coofiguration of meter statiOl1S or a differ.,..! delivery pressure from that much the City has now. Soch changes may cause the City to add staff or make inIemaI staff changes 9 > •• ":" .,' "-.---; ~~ -...:"" . resulting in additional costs. Engineer'.ng's input on this issue is key. u; RegWoIDIy Sensitivity The CPUC and the FERC are both major players in the "",-ural gas market It is ~ to n=gnize the impact of regulation on each alternative and each alternative's vulDcrability to future regulation. 1.7 A1'lIIIaIili1y The cum:nt and future avmlability of each alternative is an ~ element. This ai!Iriln also includes bottleoo:lcs associated wilh each altem:llive. For example, Southwest US. gas supplies may be plentiful, but if PG&E does not hove adequate pipeline capacity from Topock, AZ, the gas will DO! be able to reach the market 2.8 bpait"'" The utility !IIl!fi's experience in handling the various aspects of each alternative is oomido:red. 1.9 J"muidtalioo The smff handles <XlrItI:lcts, nominaIioos, and daily balancing. The impoct on administtative \\Uk is !aka! into accocnt. Consulting costs are also included in this categ<xy. If smff is unfumiliar "itlt ;odministoring an altanativC; co<!SUIting or oIher servioc cost may need to be considered 2.10 1Do....... . I ~ Cld!eNise """"n as ex:emalities in an IRP, the environrnenraI characteristics of each alt.mative will be examined in a broad sense 2.11 &lu!I'''. Each plan's elCpOSUl"e '" risk is exmnined in the final stage of the ana!ysc.. A roI:ust plan is me .... tich will still be desirable UIIdcr various advase future scmarios. This may be achie>-"Cd through the flexibility of !he alternatives. 10 3. QJaiHative Assessmmt of Alternatives First, !he gas utility undertook • broad =men! of all potential supply-side and demand-sid<; resoun:e altematives. The fuUo",ing is • qualimlive description of each of !he alternatives. 3.1 T1lIII51XIltaDoo am SlJAiy from Canadian or Domes1ic Regions Palo Alto needs to get the majority of its gas supply from 0IltSid<: the state of Otlifornia. This supply "'" come from domestic produce:; ill the SouIhwest U.s. or the Rocky MounIains or from Canada. This gas may reoch Palo Alto via several interst3le and inttasmIe pipelines. The map below is • sdlematic of the pipelines ser;ing or proposed 10 serve California. NATURAL GAS PIPE' jNg SYSTEM ~ ~ .~I~ 0_11 i \ _ JI''i7:li: .. ,,.;- _ lI''iilE' I"H"'~:)'fE 11:)("./1=10" _ ''!lIE f)ll:ll,\NQ "1 .. 1t"IN! _ \IIC .. ,II/E E,;>II~.sIC'" _ 'WC~M'E I!!X:17:HG "11l'!:"-I ... !: IIC·I~~-::"''' E~I Ci" I "'7 .. 1_ .... 0 I. 1'lII .. lOnrt,'l'IIU' "''''!:;':NCS FJg1ft Z 11 . .:07:< .. ,": :-<~~:' ~.,,;~ .. r,' 3.1.1 PG&E Al1Iematives 3.LLl Four of the five interstate gas and lransp<Jrtaiion alternatives rely on PG&E for intr.IotaIe service. M!i.ve is not schaIuled to be in operation until13le 1995, so, even if Palo Alto chooses !be Mojave alternative, a plan needs to be made for irmstate transportaticn In the PG&E sysI<m L~ !be interim. Hmve=, the following cllaracteristics of !be PG&E system must be taken into accrunt in our Img-tenn pian. Cmt "., RctWitity PG&E does not have a guaranteed rate. A discounted irIIra.ta!e rate with PG&E would have to be negOOa!ed foc any of these a!tematives to clearly be more cost­ effi:ctive than M!iave. PG&E docs have three main lines serving Palo Alto and 50 years of prov"" rdiable SCf\;"" so it is not pe=ived tc require additior.al back-up servi<e. Gas 0.." .... '" The ],(,AE optioos do allow a=ss tc the many produc= in both Canada and the Southwest. Howev.,., ecooomic tlJem' and our analvsis shm>~ that the Canadian gas pi= have and will conIinue to ti-ack liS gas pi;a:s keeping the diffi:n:ntial rela!ive~ srmlI. This t:."sponaliun option is cxttemely sensitive 10 regu1aMy turmoil PG&E future rates are UIlI:C1ain, and c';en • discounted cO<1lract approved by the CPU(' is not guaramced because: 00 Commission may be held to • p-e'Viom Canmission's decision. If the proposed M!iave Iareral is not built and bypassing the PG&E system is no longer dI1 option, the CPU(' may determine that the discount was in¥udcnt and nullify the COIlIr3Ct between Palo Alto and PG&E. 1'G&E option 1: 100% PGT Buying gas from Canada and transpcxting it ""er the PGT pipeline to California is one aItemarivc that utiIi= PG&E's inIrnstIIe pipeline system. In reoIity, MO imerstaIe pipc1ines carry gas from Canada to Califixnia, PGT and the PGT expansion pipeline. Even though the maximum rate 00. the PGT ecpansioo pipc!inc is more than the rate on the existing PGT line, statfhelieves it is reasonable to assume that capacity on both pipelines will be sold at the same price to buyers. The == of int=e capacity is the l'CO'lOIl fcc this pi .. IOiDenon. MIlly Cana<fmn produc= and marl<eters OWII capacity m the more expe11'ive PGTX pipeline, but their gas will be f<rced to 0JIJl!l"le with gas flowing 00 the less expensive existing pipc1ine. The difierer>:e in pipeline costs, therefore, will be =Iized by the gas produ= !lIIher than the end user in California. 12 3.1.1.2 .... -. The price of bot'> the e:v.isting and e.'qlallS1L'11 pipeline C3J"Iclty is unkncm" due to restIUCIllrin~ cost:; \\hich will be recovered throueh rates and the p",,,ibie FERC deCision to fOU-in the two pipelines so thai one single l1IIe awli'" to bo!h. This .ttemative is modeled as one in the IRP altlxlugh the foU""ing are three strategies that could be followed to ilnplemenr this option. i. Bid on the open market for as much POT capacity as possible. Palo Alto Ms done this as , short-tam action, but the Citv could continue "il" this strntegy into the t\Jture. • ii Negotiate with PG&E for the assignment of Palo Alto's full capacity rights on POT. iii. Rcly on gas suppliers and marl«:ter,; who have ~ on the POT expansion pipeline to supply Palo Alto -.ith its full <r partioJ gas requiI dueats. PG&E 0ptUm 2: 5Im PGTI51J% E1 Ft1so to PG&E Gr.d Under the CPUCs capacity broke!ing program, wholesale customers are permitted to reserve pipeline capacity for their core customers on the same basis as FG&E reserves fur their customers. Therefore, Palo Alto "m be required to split its ~. requiremenIs betw= the PGT and EI Paso pipelines. Palo Alto did IlOl accqx this option, so it was eliminated from fu:thcr consideration in the analysis, To take ~ of the res:rvation righls olfm:d by the CPUC, the City "oold have paid the highest ~ l1IIe charged on eacl1 of the existing systemS. In the ca<e ofJ>(,"T the me, '" least for the interim. is low. Although, a puceeding cum:nIly at the FERC may roD in the cost of the PG1X resulting in a melded rare "Nch is much hig!H:r. Capacity on El Paso .ls in = and is selling on the open market for less than the ma>.:irIwm ra!e. Even 1bough Palo Alto WOldd be saddled with fuD-pricW EI Paso capacity, this <¢on could be b::tter fiom • 'IRC stand point than taking 100% of the capacity CII EI Paso bo:ause the POT pmim will be used to .. ;cess less ~\ .. C2nadian gas supplies. Stil~ a capacity giut exists, and it was not m:ommended that Palo Alto take 00 the responstbility of the pre­ arr.mgal deal. This option 'MIS IIlOO: costly than 100';, POT and also required a ~ corntnitrnont which limited our f1=bility to chose other optiom. 13 3.L1.3 3.1-1.4 3.1.2 PGAE q.tion 3: jlJO% EJI'rIso to l'GM's Grid El Paso c:apacirj is cumntly available on the """" marXet It is in over­ supply and is being soid at a diswunt. This alternative is expected to have a high TRC because the price of Southwest US gas is predicted to be slig/ltly higher than Canadian StJjlpIies, but this alternative gives the City greater t1<X1bility in the future since Palo pJto = choose the term of ccmmitment. PGAE q,t;on 4; eo... Sd!&:tiption Un!l11991, Palo Alto was a core s<IIJ"aiber of PG&E. Core Subscription service includes gas supply and gas tr.lIlspll,1ation This option is still avaiIJlble to the Citv. This aItemaIive does not afford Palo AIro onv of the benefits of discooined il!te:state capacity ,; gilts. In addition, IIiis plan renders our entire gas suWI, and lrn!l:;prI!Alion subject to CPUC regulatioc. M:ljave could take the place ofPG&Es inIras!ate transportatioo service which the City poes<n!Iy d<pends on. Mojave will operare as an inI=ta!e pipeline bypassing PG&E's sysIfm to connett CUSIOIner5 thereby avoiding California Public Utilities Coolmission jurisdietiO!t E>en!hough capacily brok<ring has caused Palo Alto to DDke an immediare decision among the PG&E alternatives, the decision about ~~ave is imminent. Cmt ad IW"udiljlY Mojave is offering a IS-year guar.mtced transmission rate of SO.47 per d=therm (D'IH) at a max:imum of 1% escalmion per year ,.hich the CPUC has no corttroI OVCI". In!C<StIIe capacity on either EI Paso or Tran.western. will also be required, and IaIc5 on these pipeIiDes are expected to be disaxm!c:I in the future. Since Mljave is ooIy one pipeline, as opposed to PG&E's three, back-up p-oIeCtion may be required. R.:seard! bas sbown that ov..-90% of the municipll gas Wlities in !he oountty =eM: gas fium • single pipelino !hal CX!CJds an ~ distance of 40 rm1es to an inI=tate pipeline Nevertheless. if Palo Alto chooses this option, provisions will be oegotiB!ed to miti@llle against ~"Y or CUI"I3iJment conditions and to OSSIR gas deliveries for our scasonaI residential load. Some capital C<JSt will be incutred to modify one or more of the City's meter 14 · , , :7 .... <.- stations to krtm:onnect \\.irh MOj2ve. Engineering has completed a hydraulic study "hich shows LIJat one receipt point is not enough for the Cir:'s disttibution system to functi<hL Palo Alto's Engineering sWf r.as also <le\'elopcd a propos<:d inIm:oonect configuration and associated cost estimates. The responsibility for the cost of CXIra stations "iIt be part of the negotiatioo process. Gas O!mmocitr Many producer> are accessible via the Mojave pipeline. The most omious so= of supply is the P=rian Basin ",trich can be accessed by the El Paso and Tr.lI1S\\estem pipelines, The price of S<:ruthwest gas relative to Canadlan gas was a criticai variable in the IF-I', The variable was studied using indu.<;try experts to == the probability disttibution of the po<iSible future discounts for Canadian gas. Sel1ioo 4,3 discuss<s the probabilistic phase of the analysis. and Appendix C contair-.s the probability distribution fo< the COTIl!nodi!y prices. QJmiiqnal Coomns The main ~oml concern is how well the CitYs distribution system = function with only one receipt point rather than the current four receipt points, As discussed above, the City of Palo Alto Engineering staff has modeled the hydraulic effects on the system of an intm:onnect with Mojave and has fuund that multiple <:mno:ttim points wili be needed Additional pip<ng and meta stations will result in a higher investnlell! cost associated wim the Mojave alternative, RcgnIaOpQr SmWivi\V This !I3nsport3tion option is relativdy insemillve to reguIat<X)' rurmoil Since Mojave will be an inrerstme pipeline, it "iil be subject mil' to FERC regulation The prqiect will be built under a special e:qx:dited c'frtificate wtrich means that the M!jave ~ must oo.r any cost ov=, Therefore, the rate is guanuneed f<y 15 years, and Mojave can !lOt impose any i3Ce increases regardless of the final C05t of the J:roject or any 0!fJer factors, Aywla1J!jty Mojave faces considerable <WOSition from PG&E and the CPUC The utilities staff has r=ived varying opinions from gas industry c=>Ultanti ilDd mar'Aetm ranging from a 50% to 100% probability that Mojave will be built The probability of. lateral to Palo Alto is also UiJlclrnm 3.2 OwneIsli.p of Califoma ProWctiOD Another interesting optioo the City has is to participate in a partn<.nhip to drill ga'i wells in California and adlIa!ly ""n its ga'i supply, It is unlikely that Palo Alto would ;", able to, or wan! to, rely solely on its O"'ll wells because th<re is a high risk assocU!Ie<l with 15 • l , :-\-: "'~'-; -':~'~ mlling b<olusc of the huge uncertainlY about the: level production and the life for any givt:n \Will. However, this option may supplemen! oot-of-swe gas supplies and is included in the !'1'5 IRP . Vemda National, a California gas producing company in Sacramcnto, ha. discussed porIlItrSbip amJIlg<met1!S for reserves and drilling progrmns. This is the only oppor1llnity tba! exists b!he City at this time. DSM g<n<:<aIly refers to comervatioo measun:s Mlich the customers of !he utility undertake OIl their homes or businesses. Tbc:se measures result in changing the demand paItml kr gas. In this "")', 1hese OSM resources an: often refim:d to as an altemative soon:e of supply. Ukc QIIif.:.rrria gIllS, this optioo docs not relieve the City from choosing one of the UiiiSjlttl!lli.Jn altmJatives di"''5sed in this report. Om DSM is included in the 00, baa"",·er, a JIlIJJd.<:ommodity OSM srudy is urxJern.ay ",hich will jlI'O\ide more iufoona!im about the poteotiaI kr DSM programs in Palo Alto. 16 ' .... ---- 4. Decision Analysis Making resoI.:I'CC planning deci<iOl'.s is a complex problem sin«: there are usually many competing supply-ar.d demand-side resouroe5. In addition, th= decisions must be made in spite of the fuct that rmny future events carmot be kr"",n or predicted. D:ci;ion analysis techniques provide a systematic way to break dov.n complex decisions into manageable co!!1pOflaltS. Each stage of the process gives anal)~ and decision-makers valuable incite into the problem and helps to focus on the most critical elements surrounaing the decisioo Ihm must be made. The steps in the decision analysis process are sl"mn in the fon""ing diagram: Assess Devel~ Evaluate Decide Establish Si:wticn Alternatives' Risks Among an Action lI-.famarion Alternatives Plan Problem D:!=inistic Probabilistic Evaluation Action Fonnulation Analysis Analysis Communication 4.1 Prolllem Fonmlalion The puzpose of the problem fomruiation stage is to define the problem and the set the scope of the analysis. The steps include: (I) state the problem, (2) genernte a list of aItematives and ur=tain ,..nabios, (3) designate the criterion or criteria upon which candida!e plans will be compared. The quostioo "hich prompted the lRP was: "Should Palo Alto make a IS-year COOhllmt .. It to Mojave pipeiine"" In order to ar.swc:r this que;tioo, it was necessary to analyze all alternatives and choose the optimIlm plan for the IS-year planning horizon. In addition, the City had to chose whether or llOI to """"i'I a prearranged deal with PG&E foc interstate pipelille capacity. T"" criteria were used to examine eachpJan, IRC and rnte impact TRC includes the resource cos-., tr"d!lSJXI!l'Ition cost. the cost of demand-side progrlt!IIS to the utility and to the =tomer.;, and the environmcmaI cost to society from tuning !IOIUr.!l gas. 4.1.1 AItomaIiva A list of all the feasible alternative for meeting the CitYs gas supply needs was genc:ra!td. The alternatives were discussed qualitatively in Cluq1ll:r 3. The three decisions and all the alternatives are sbown in Table I below. 17 SurnmaryofAltemalives lnlersI3le Gas & Tra"spJCwion Mojave PGT EI Paso PGTIEI Paso Con: Subscription Ownership of CaIifumia ProduaiOll Yes, do j21icipa!e No, do 1101 pattkipate Demmxl-Side Mana"",""" Nooe - Levell Level 2 l..eveI 3 Commens Domestic gas to El Paso to Mojave Canodiz gas 10 PGT to PG&E I Domestic gas to El Paso to PG&E 50"10 on fui, 5i)% ll!I El Paso to PG&E Buy aD gas and ttlIilSpJltatioo from PG&E I Do or don~ patticipare in a Califumia I ~ prosram for a patiOll of the C"Ys load I 15% of demand by 1008 1 10% of demand bv 2008 [50/. of demand by 200S Taele I Tho:se alternatives ""'. also be illllS1I3lci in !he form of a "decision tree" "hich is sbov.TI below in Ftgure 3. Eacb paih from left to rigfu !hrough the _ ~ a plao. lhis combination of aItoroatives yields 40 plans. --- 18 4.1.2 4.1.2.1 4.1.2.2 4.1.2.4 UDl>!ftlin Variables The uncertain variables thai affect the TRe and rate impact for the various plans fall into the ge:=aI categories 'xlow. For the most part, each ,,!Iiable was given a nominal, hi gh, and low value. and these range. are described and tabulated in rmre detail in Appendix A GasDmrni The demand was set at • base forecast for 1993 and then escalated at an uncertain rate. This variable growth rate encornpasscs the possible growth or e::onomic decline or S'jstern b)pass by Palo Alto's indusIrial customers that the City could experien.. ... v.tich wooId change the amount of gas consumed over the 15 year period. bjldion _ Discow Rats The inflation rate is the average rate of inflation over the evaluation period. The discount rate i, 3% higher than the inflation !lite. GasP1i= The price of gas in the Pennian basin is the index for the price of all the gas v.+Jch reaches CaJifOO1.ia. The escalation rate aver the planning horizon was treated a.; an Il!lIX<1ain variable and the discount or premium for Canadian supplies \Va.; also an un.~. PG&Es weighted average cost of gas (WACOG) was a calculated value from the resulting prio::s of domestic and Canadian supplies. Pfpe/inI! eo. .. ,itJ ...... Levd md Rat!S The batiSjXJI tatioo rate fa-every inttastate and int<rstate pipeline in the analysis wa<; a variable. These include, Mljave, PG&E, PGT and El Paso pipelines. An additiooai cost is added 011 to all alternativ<s that include PG&E ttansportation. 11-""sttanded oost" is <lui: to the restructuring of some PG&E =acts, and all ~ .. ill p"y a portion of them. It was assumed thai pipeline capa:ity "ill be reserved to meet Palo .AJto's cold wealher f<:.m:ast less any DSM It was also assumed that surplus capacity could be laid off at a fu!ction of its cost The high value of each layoffw.able was lOO'V. of!he original price paid. This all""" us to test the value of reserving eJCBCtIy as much capa:ity as necded. The model also includes a cost that is incurred by =ving too little capacity and having to buy defickncies at a pn:miuru. . ~ The amount of st~ avaHable and the cost of thai storage for each 1Iat1Sp(X!a!ion a11<r!'.at.ive was <:oosidercd. Stornge affects the cost of gas 19 4.1.2.6 4.1-2.7 . : ~'-" .':"-.' '"': ... . ' ,'.-' --".-,.~ . be=Jse it allo". more gas to be purchased in the sumJl1fr months "bile prices are lower. Stor.g< on PG&E s system is charged based on total througf1put It was assumed that slDr'3"oe an lvloj .... e will be charged baled on 5101'age capaci:y although the oominaJ """" 3SSUllle5 no 510rnge is 3V'"oilable on the Mojave system Other Moj;1¥' CasIs Back-up co;! -This includes the cost of back-up ..moe fium PG&E (l{ other means of JlfO'iding servi« in the case of an emergency or curtailment Gas pipelille systemS in g:=al are vay reliable systemS gina: they cootain no mechanical parts, and stafl's """"""". has sllmm :bat over 90% of the municipaJities in the CO\JIIlIy are "",'<XI by a single pipeline. ~mbdess, this cost was added as a precautioo to • poosible pclicy decisioo requiring back-up to a.;:munt for the fuel: that Mojave will have only one pipeline serving Palo Aho v.hile FG&:E has three. A reliability study is und"",.y by Mt~"''e staff members v.hich will be used to make the decisioo aIxxtt the ne<d lOr back-up sa-vice. Palo Alto's rnpital cost to tie into the Moj .... e 5)= "as also considered to be an UIlCCSIain variable. DSM DSM IIDgrarn> are conservatim and load managetrent progtamS that impoct sales and peak day usage. Three DSM levels were evaluated in the IRP. These levels provide • c"",p<tI .. .,ive rnngo of ,"1llues that model ~ of llSM IIDgrarns that are similar to tlIcse IIDgrartlS currently being implemented by PG&E. The variables in the DSM levels included DSM Pe:fonnance as a % of 1Blgeted DSM and the D&"l enviromnenIal benefit. The environrnmIal be:ldit of the avoidance ofburning gas was calculated using the California En«gy C<llnmission's value of S33/ron of Co, = $1.I9/MCF (thousand cubic feet.) Tho level of DSM targe<ed, the utility program COSIs, and the CUSlortrr coots wore set ,'3Iues, IlO! variables. Theses values are des<n'bed beicw. An ~on V;IIS made that the program would begin in 1995 and cooiinuc: 1hroughoot the ewluatioo period • Three levels of DSM impoct 10 be acccrnpIished by the year 2008 were assum<d 100el 2 was 10"10 of load lOr the year 2008. The high level was 15% by 2008, and the low level was 5% by that year. 4.1.1.8 • Annual participant costs llSSOciatcd v.ith eac.~ !evel of DSM are $220,000 for the nominal c&Se, 5750,000 for the high case:, and $41,000 for the tow case *' Annual program costs 3ssociated ~th each level were S110,OOO for the nominal case, $290,000 for the high and S41,000 for the low case The Domina] case assumes that .all program:; are implemented with full participation. In general .these programs are mosdy water hearing., space heating, and process -:.ontrols rebate programs The low level scenario was obtained b)' selecting the least expensive measures oul of the ~ominal case, which iiIe mostly residenbaJ, commercial, and industrial water heating needs. The high leveL scenario is comprised of a pro,gram sort of expensive measures that are added 10 the nominal ,ase, Typically. these measures were remdential space beating, and comrnercialfindustri al space heating and process controls. DriIIiIfg The production rate was set at about 1 {'II/a of Palo Alto's demand tn 1994. The investment cost to achieve this level of production, the ongoing operating costs and the dec] ine rate of the wells were treated as uncertain variables. Data for this analysis originated from an enensive report OD California Production done by bean Research corporation in 1990. The values were escalated 10 current dollars and compared against 3I1. .aauaJ drilling programs which Palo Alto has re-.;ewed within the past two years. Rquktory .t Alimini,'''.:'' Administrative costs and costs associa..ted Vl.1th participating in the FERC and CPUC proceedings were includ~d in the model. The model assigns the low, nom.inal, or high value to specific-alternatives_ PGT, EI Paso, and Moj ave all have jow a.dministrati. 'Ie costs, 50% PGTlE: Paso and drilling have high adm.inistrative costs, and core subscription has a low adminiMrative cost The administrative cost associated with DSM is included in the utility program cost The role of these uncerta.in variabT es in 0 U-! analysis may b: expressed in an influence diagram, Figun:. 4, 21 1_-' -' -=:..! 22 4.2 DEImMlNJSIIC ,,,,""AL YSIS 4.2.1 PIaplse The JlIIIP05" of the detmninisti< phase is to develop a value model ,.ruch links !be IDlCCI1ainties to an outcome so thai the all<rnatives can be CO!llpOled =l to reduce !be number ofvariables L~ oro.:-to fOCl!S attention on only those variables that affect the decision being made. In Jcne the UAC "3S prESC1l!ed with the results of the dc!er.ninistic analysis. However, enhancemerrts to the model required • revised deterministic anal)!sis to be presented in September. Sensi1Mty AnaIyn. The 'nominal plan' is the optimal policy (the one \\ith the lowest IRC or rate impact) when all variables are set to their nominal or base valu<:s. The oominal plan using IRC "'as to purchase dom:stic gas, transport it on Mojave pipcline, participate in California drilling, and do level 2 DSM. The nominal plan using rate impact did nO! include any DSM These optirnaI policies did DO! chmged fiom the original de!enninistic analysis. The nominal plan is !JOt a Lccoonnended action plan. It is mly • base line from which II> test !be atrect of the Il!1I%l1ain variables on the decisioos being made. It is also • base line from ..nkh to test the affect of a decision 011 the other decisioos. It is impxtam not to focus oc the nominal plan itself but to focus on the variables and altemati .. es that cause the plan to change. The sensitivitY of each L.LIlCfI1Bin variable was tested by rurming the model with 1be target variable at its high and low values while holding all other ,wiables at their nominal value. The ,'Brlable was deemed sensitive if the optimal policy changed from the nominal plan "hen the target variable was set at its high or low c:x!reme. Tables 2 and 3. below, show the change; from the nominal plan at the low and/or high ,-alues for each sensitiv'e variable. Blank spaces indica!e that the chasen "ltemative was identical to that chosen in the nominal plan. 23 Olange in Nominal Policy at Extreme Values of Uncertain V2riables TRe Variable SIsre IntersIa!e Ga<; O .. netShip of & California Tr.mspottatioo I Production NominAI PIan Mojave Yes Canadian Ga<; lY.scount -low PGT Mojave Back-up cost -high PGT IXilling Investmenl -high No IXilling CFating Cost -high No DSM PertixlllllUX -low DSM F.nviroomcIlIaI Benefit -low Ie O.ange in NorninaI Policy at Extreme Values of Uncertain Variables RA.TE IMPACT Variable State lntcrslaIe Gas o..nership of &. California Transportalioo Productioo NOmina! PIan Mojave I Yes Canadian Gas Di=unt -low PGT M:Jjave BId-up cost -high PGT Drilling Inve:;tm,m -high No OOIing ~ Cost -high No Southwest Gas Price -low No Gas Use -bigh e 24 · , DSM Level 2 Level I DSM Nooe Leve! I ~ .... , -,1;,,-:) 4.2.3 ~c Qsmme oad R&c<o" ....... jo ... lhe critical decision at this time: is to choose the bc:sI ~ 1I".mspct laticn altemariv:s. ~ other 1\\0 decisions, drilling.oo DSM _ no! depolObs 00 the trarISpON!ion dccisiOlL Th=t"= the drilling and DSM clccisions .... no! considered in the remaindo-of this analysis. AMthcr """"" far JlOS!PC'DIlg the DSM decIsion 31 this ti,,,. is the lWIti, coI!lIlXXlirv DSM stIJdy "lll<:h is cm=lv undc-.\2'. ~ SIlldI.' ><ill cIarit\' the ' DSM potOmial for Paio .-'Jt.o, Also. it is' befie\'ed ihaI the drilling decisiai .. ill need 10 be eva<UUd on the basis of! specific projea or ~ Rt:sul!s from the integrated DSM ,rudy and an: ~.fc drilling projOClS sbooid be incbMW in the next gas IRP. Since the probabilistic phase is an analysis of the gas tr.IrISpc<tatia crcision C<!/y, the numbe:' of plans is roduced from 40 to five ~ dc:ision tree for the probabilistic phase ;s shnv.n in Figure 5. Several ,-.riables W<>'t fOlD! to be scm-itive fur this dccisim and ~. ~ disclIs...,J in the !lClC! 5CCticn AJso. since DSM is the only decision '.,n;<:h affectS r.!le 1mpacr dii!=!)' than it affeas 1RC, ally the IRe anai~,is "-" done in the probabili<tic phase. inlersta\& Gas & T ratl$portation GASIRP DECISION IRE" ProcabiHslic Pha •• ?G"i FIgIft S < Tr;.C ., \ ". ,,/ -.;y.J 43 PROBABIIlSTIC ANALYSIS 4.3..1 lUplse ';he purpose of the probabilistic phase is to closely examine the boundaries of the uoo:nain variables, inteniew ""P"rtS to constItIct the probability distnbution curve for each critical variable, and run the model using the probabilistic data bpett ~ and &y Voriables It was d<tmnined that the variables from the deterministic phase could be divided into two~: (1) ur=tain variables whose values could DOl be koo"n prior to decision making time, and (2) values tbot could be negotiated, and therefore, known at the time • decisioo is made. Fer example, the COl'! of back-up service for the Mljave aitemative was det<nnincd to be a negotbble ,mue rather that .... uoo:nain variilhle. The variable was set to its low value for the probabilistic analysis, but it may be incroased iaaementaUy to <kI:mnine Palo Alto's breake ... "" point mth regard to this cost. Also, • new negotiable 'mile was added, the daIly demand quantity 00 Mljave. We have been negotiati"g mth ~ave and performing an c¢mi2ation study to dctcrotine the most cost-effective maximum daily demand quantity foc the conIraCl Daily and moothly p:naIties foc being OO! of balance are the main concern. The goo) is for Polo Aha to meet its peak requiremeots "itile keeping <XlSls <Im>D. Three variables wo", 1reat<d proOObilistica!ly: (I) the discoont fur Canadian gas supplies, (2) the cost of Sou!hwest US gas supplies, (3) the market price of EI Paso pipeline capacity. The price of dorr.estic gas and the relative price of Qmadi"" gas have many possible future \ahles. Two official price fun:casts of C.anadiM and Southwest US gas "ere orouned. The tim was the Cambr'4e Energy Reseord, Assodatioo's forecast "hile the ocher was done by McGraw Hill v.hich "lIS provided to Palo Alto by PG&E. In additioo. a gas industry and economics expert "as interviewed using a systematic procedure for assessing proilebility data for 00Ih ur=tain gas price variables. The Canadian gas price variable was assessed a second time by IiIIOIh<r industty expert to verify the results. LikeMse, the expert was inl=iewed to assess his judgemaIt about the third urx:enain ,1Iliab/e, the markc! price ofE! Paso pipeline capacity. The resulting probability distn'butions are s~ in Appendix C. E.'<IraCting the c:>qJC<ts ~.1 about thc variables is • rompkx problem Each interview IasIed for a minimum of twO boors, and a huge amoun! of L-tfoonation regarding the ~5 views about the relative likelihoods of future scenarios W;t; extracted. 26 , . , However, some basic deductions can be made about the experts belidi; by .xamining the distributions. rOt i."1Stln<:e, ~.,. e.'qlCI"t beiieves that it is likely that Canadian gas wiJl be priced relatively I"""," !han Soothwest gas supplies. The basis for this is the finding and produ.."tion = in Canadian gas fields "'" lower and it is pe=ived that the gas consumption market in Califory.ja will continue to be vital to Canada. The probability distribution fc.r the price of Southwest gas indicates a beiief that the price is just as likely to increase as de=ase. The assumption which leads to this conclusion is the for=; that woold lead to a price increase (growth in US gas demand) are coonI=ea by forces that would cause • price decrease (tecI:nologicaI i~ ""wnsizing of US oil and gas coo;panies, and " more rompe!itive California merl<e! place.) 11"" probability distribution ior the)Tice orEl Paso pipeline capacity indicates that the expert beli",'es int..'rSUle pipeline capacity to Oliifomia is in excess supply, and that capacity is I ike!y !() be heavily discounted 00 the open rnaJi<et as a n:sull ~ The results of the SlUdies and the int"",i""", ."""" input to the roodcl, and the altematives were evaluated again The PGT and Mojave alt<marives Wt:re both atttactive with regard I<l TRe v.1ten the rest of the variables were set to thcir norninaI values. The sensitivity of each neg<Jliable varillble '''''s tested again by nmnlng the model with the probabilistic data for the C=dian discilUnl, price of Southwest gas, and price of EI Pa", Capacity. A variable "'''' deemed sensitive v.I!en the optimal policy changed from the nominal plan as the target variable was set at its 0lCImIle high or low wlue. Table 4 bel",,' shows the """"irive variables. 27 NEGOTIABLE v ARlABLES -PRaiABIUSlIC PHASE Variable State Alternative ~=----------=~~~-------------~ Maximum Daily QJar.tity Moi:< ... e -~jgh PGT Mojave Capi::a.l Coot -high POT PGT Table 4 As m<mioncd above, the maximum daily quantity, Palo Ahds capital cos: to inten::onn«:t with Mljave, and back-up cos: may he able to be pinned doml by COOII3Cl negotiation. Sioce PGT and Mojave ~ equally aIIracti-;e alternatives, the final recommondatioo will depend O!l the n:sults of Palo Alto's negoIiatioos v.ith PG&E and Mojave, Fipcs C-7 and C-8 in, Appendix C show !he cos: components for !he Mojave and PGT alb:matives 00 an annual basis over the IS-year planning horizon. Figtae C-9 is • tctalized amwaI cost ~ beOO:en !he two alternatives. TlIese graphs d:pkt expected vaJues only; the risk involved for each a1temallve is represented by Figure C-4. 28 5. RECa\1MENDATIONS So I Short-Tenn Action Plan Two shM-lerm actions were recommended and taken as a result of this lRP. First, the Cily rejected !f-", pre-arranged assignment of capacity by PG&£ This action did have Ioog-tenn irnpfications since one at the five alternatives \\35 el:minated Secood, the City bid for base load capacity on the PGT pipeline fur a lerm of II months beginning aug\5t I, 1994 (the <:xl of our current gas supply coo!ract) and ending June 30, 1995. lNs will secure inexpensive inr=tate transportation for Palo .. Jto fur the interim period I".uo Alto will still have til make arrang<rn<n!S ,.ith • PGlX shipper t<> supply gas for the peak oc:eds over the capacity the City 0\>llS 011 PGT. After June 30, 1994, Palo pdt<> may P"'P"'" another inIerim strategy before service <JIl MJjave commences or participate in onoIher sho!t-or long-term w!lmribllelll on the PGT pipeline oc enIer inro an arrangement ",ith a PGTX shipper. 5.2 long-Tenn Action Plan The PGT and lvlojave alternatives are equally attractive Icng-teml plans because the IRC !Or both plans are nearly idell!ical. The final rtwtttt,,,,,idation will depend on the =1s of the oegotiaIions currently being conducted "ith represen!3tives from lvlojave and PG&E. 29 APPEl'IIllIX A Description of UnceJ1ain Variables In order to UIlderstand the c!a;i,ion analysis p!"Dbierr. the altem2li\es as weI! as the un=tainties ..mch affe<:t the aItemati"", = be identified. This is part of the problem fomnllation phase of the process. In this apperxlix. each ,..nable ts described in deW!. A.I Gas Demaod Demand, for p!llp<lSfS of the IRP. was tre:lIed quite diffen:ruly than in the City's gas use furecast. To forecast, staff uses an economdric model "ith a SIlIndard deviatioo of two in order to bound the forecast ,,;!h a high and low rang<. Since the objective in the problem fhrnrularioo stage is to understand our r:mge of uncertairuy rather than to fore=! or rmlict the future, this method was no! used. Instead, an average forecast for 1993 "~ used ,.bile the annual growth rn1e was coosiden:d to be an ~. This approach sirnpli1ies the derr.and in the model and can simulate the Iong-tmn growth or economic decline pos5lble fur Palo Alto. Grov.th may be caused by an increase in ecooomic activity in the City wlule a decline may occur as it result of business le:!ving Palo !I.ito or b)passing the gas utili!)'. The historical usage and !he scenarios that \"",! tested in the decision analysis :Ire shO\'oD in Fig!ire A-I. 700- 5<IJ- "" - City of Palo ,lJ!o Historic Usage and FtJlrJre Sanarioo -.; C\JTl.iJIIE s..:!!'iL'tlCS -------- _-...o---,-----~ --------------------- • ~--------------- -MIS':"C~CAL-:-~ -~ --0'!'i,1'oIO :~':'SiE OE-:RE,.I,SE WCi'l'E.'SC FIgure A-I A-I '-,. OR OE'~:Sc -. ~'. A.2 FIaancilII Rates A21 Ir(laion Rae The infIaIion rale ";IS Irea!ed as an uncertainly "ith • oominaI value of 4% A.2.2 IJiscoua Rill! The discount rate was tied to the inflation rme; it ";IS calculated as the inllatim rate plus 3% A.3 GIIII PIicos A.3.1 Sot8IrwesI US c... Prices The price of gas in the Permian &;in is the key factor for deImnining the price of gas 10 ('.alifomia. A IlOmi11lll value for 1993 prices was used "ith a variable escalation rate, tied to irulation, over 15 y=. This variable was treated was given a probability distributioo in the probabilistic JX!ase lOr the analysis. A.ll ~ Ga> Pricer Canadian gas prire; ~ for the CaIifcmia marl<et place are indexed 10 the P<rmian price. The uncertainly in the model was Ihe discount or premium for Canadian gas relative to P<mlian gas over the IS year period. This uncertainty was also further analyzed in the probabilistic phase. am, SItJscripIiofr Gas sold 10 core subscription custorne!l; was nxxleIed as half at the ~ Jri:es and half at Carmdian prices plus an :Idmini..<;lJ:ative fee escalated at the !lite of intlaIim. Seawtd Dijfdtl/Ui The differcx:e l>c:I:w= summer and winter prices was set at $0.20 per dIh. The diffen:nce was estimated by """",ing historical data m gas prices in the PInJIian basin. A.4 Pipeline Rates A.4..1 PG&E The variable iepie,c"ti:lg PG&E inIrastate trnnsporta:ion rnICS was given a nrcnjnal low, and bigb value. The escalarioo rate over IS y= was also a variable with the nominal value "f"'l to inIlatioo. The nallspaLatioo rate was multiplied by the actual usage for each year to srrivc: at a tcIal oost for inIras1ate ~ There is no capacity layoff; the forecast is revised periodically. PG&E will be assigned 3 portion of the Canadian gas oontract resttucturing costs. "/bese COSiS will be passed m to customers. The money will be paid by direct bill IlItber IhIIl a surcharge of a period of three years. A nominal, low, and bigh value A-2 ------- , .' , for this charge 'v.JS a5,;1Imed based on estim,lled total bU)'-<l1l! casts of 5150,000,000 10 $300,000,000. A.4.2 Mojave The MojaYe transportation rate was treatc:d as a variable in the deterministic phase. The nominal and high rate """ $().4 7 per dth escalated at I % per year. The low range """ SO.4O per dth based on the PffiSIbility of a Iowa approved rate by the FERC. A load fJJctor of 95% was used to determine the lola! cost of lldl6jXli attion on the Mojave system The effect of laying off capacity on the Mojave system """ tested by setting the value of laid off capacity between 0% and lOO'A; but the final analysis assumed ~1J.lt ZlOIJt of the Mojave capacity would be laid off A.4.3 PGT T"" main factoo; affect the PGT rate: roll-in and restructuring casts. Roll-in is a term used to descnbe the process of combining to. low cost of the existing, old, clepreciared PGT pip>1ine ,mil to. vigil cost of the new expansion pipeline, PGTX The result is ident;ca1 rates for service on either pipeline. The oppo6iti: of rolled­ in is incremenlBl. Restrw:turing costs will be added as a suzcharge to the PGT rate as a resuJt of the conttact negotiations between PG&E and Omadian gas produa:rs. The amount of the cootmct huy-oot "ill affect the amoum of the surcharge b=use it will be divided up amoog all the CIIStot=-s on the system Three distinct cases were chosel for d., nominaL iow and high IlIIlge5 for the PGT rate. The nominai case assumed a modernte restru::turing cost, $225,000,000 for the entire system, and roll-in by 1995. The low case assumed a low restnx:turing cost, $150,000,000 and no rcll-in wlule the high case a'iSUrI1<d a ~igh restructuring cost, $300,000,000, and immediate roll-in. As for the amount of capacity purchased, an aOO:jua!e level of capacity to rover Palo Alto's winter peak was assumed. Unused capacity greater than 10'10 ova actual usage """ assumed 10 be laid off at nominal, high, and low vaJues of [00'/., 125'/~ and 50"10 of the price paid A.4.4 8ABo Pipeline capacity on the EI Faso NaturG! Ga& pipeli!!e is av,flable om ell open nmI<et. The high value for the price of this capacity WI!S assumed to be the fuD as-bilIed rate whiI. the nominal value assumed " modera!e discount and the low value assumed a o.q.,. dis<ounL Like PGT capacity, an adequate level of capacity to cover Palo AIto's winter peak was assumed to be purehased. Unused capacity greater than 10% over actual usage was assumed 10 be Ia.~ off at nominal, high, and low values of 100'/" 125%, and 50% oethe price paid A-3 "'-. ------. A.5 SIotIge .u.1 PG&£ Even !hough a proceeding is unden;llY \0 unOOndIe staage from the existing trallspu:1lIIion rnie, Palo AlIO will still be required 10 take some of PG&E's storage in or;J". to rc:Iain our peak day protecticn status, The amount and the cost were 0<pI!rlIIe variables. The nominal case was the amJ\.."1I of storage prue Alto should be allocmed today at the CIlITeflt cos:. High and low values for both amount and price wm: a1so rrodeled. A.5.2 M<ifrw~ The nominal case assumes !hat 00 stornge will be available OIl the Mojave syBt_ however. a bigh range was CS'.ablished equal 10 the amoont of storage currently <IVlIilabi. (JQ the PG&E syS!errc A ""'go of costs were rrodeied. These costs wm: assumal 10 apply only to the amount of gas injCCled or "ithdrawn limn storage r:Jther than on the \OIal sy= throughput as the PG&E storage is modeled. A.6 Other MljIM a..t. A.6.1 Cif1i1d Cos! The capital 0JSt for Palo Abo to inb;JOOInltX! with Mojave pipeline is a function of the hydnmlic requlleWeUlS of the gty's distnOOticn system. The nominal, low, and high values fer these costs were derived from Engineering's pre1imina:y cost estimates based on various scenarios .... hich retj'.rire different numbers of meier S!l!Iim 10 be installed. The costs occur in 1995 in the model A.6.2 Ba:k.1fJ Cast A.7 DSM It was assnned !hat su:ne fonn of back-up ,"",ice, presumably from PG&E, would be subsaibed 10 as a precautioo against Mojave's one line being intmupIed. Since PG&E does not offir back-up =i....., at 1his time, a 1>'i<!e range ofpossible V"d1ues wm: establisbed. Reliability studies of the ~1:Jjave system may render back-up from PG&E ur.r>e<essaIi'. Another foon of back-up being coosiden:d eOlllils Palo Alto contracting wilh G!her Mojave C\lStOI1Im Ie assign pipeline capacity to the City in • period of cururiimfm. These costs were esVmatet! by having infonnal convt=tions ..;th the Mojave shippers v.no have t!Xj» I ed interest in such an arrangement. The cost fur the capacity =igmnmt is estimattd to !3lI well within the variable !3I'lges. A. i.l EladN4"ZJ~ Bm!!fit The nominal enviImmenta1 benefit was calculated usb!! the California energy Commission's wiue of 533 per ton of carbon dioxide wbidl yields a value of 51.91 per M:F of natur.aI gas burned. In the mode~ every M:F of DSM was given '"' ~ credit of SI.91 because the Jroduction of CO2 would be avoided. The efre<t of changing this V'.n.vle over • range of $0 10 55 per MCF was IA:sIed. A4 · ; -i_ : . . ~ ...• ~. "'<-~~~ .. .. '>'_ " i ,-~ A. i.2 Peifomsro? The cost to the utility and to the CtJStomer were set in L'>is analysis. The ,ariable was the pt'lformance oi the ~ in O!hcr words, how much DSM Palo Alto actually r<nliztd for the same amount of money. This value had a range for 50% to 15Q"/. of the ~ amount of DSM. A.8 Califo ..... DrilliIIII A.S-l Jnv~ Cost, ~ Cost, and Dffline Il4e The productioo rate was set aI about 10'% of demand vmich the cost to achieve this amount of B3S was IJ!1<l:rtairL The investment, oper.tting rosts, and decline rate wc:r: obIained using R= Researd!'s 1990 report on Califcmia production and Gala received from • proWcer in the Sacramento basin "00 was proposing just such a~ in 1992. A.9 ~ and AdIDaisIraIiv. A9..1 AdmbGIJai~~ The admirlistnIIive = of each aitemative were originally treated as variables. ~er. the 6naJ ana!ysis as&mJfX! sct administtative costs for each aI1emative. Mojave. PGT, and EI Paso were assumed to bave nominal administrative rosts, \Wile the 5(1'50 split had a high cost and the core subscription a/tematiYe had a low administraIive cost The ""'ges were based on the "lIue of staff time need to administer each aJtanatiYe. A.9.2 FERC All altematives included a cost for lobbying and litigation at the FERC. The amount was treated as an uncertainty. The values include staff and con:.-uJting time to deaJ with me issues. A.9.3 CPfJC QUy alternatives v.ruch involved PG&E inttastate =pottation service included this uncertain variable. The nominal case was based on the current CO!l.>.uting budget and stafl" time spent 00 CPUC f'OC1'edings and issues. A high and low boundary were assumed. A-5 DEIlR'\1INISIlC PHASE UNCERTAIN V ARlABIE RA."I'GES • IXIS Value Value Value /. /, -~?o ,. ,0 lnflation Rate 4% 2%~ 8% SouIhwcst !!"S csalatioo % 1% .} % 4% Cf"Iadian discount % 0% -2QP1o 20"10 • Mojave Rate 0.47 , 040 0.47 • PG&E Rate 0.68 0.42 1.00 • PGT Ra!e 0.17 I 0.15 0.31 • EI Paso Rare 027 0.16 0.39 LayolfValue % of original prire Mojave Layoff ~'O 1000/, PGT Layoff 1<Xn"O SO"I0 125% EI Paso Layoff 100% SO"I0 125% "Stranded PG&E Cost Slyw 44,000 29,200 58,400 PG&EAmount mc£i\'ear 300,000 189,000 400,000 • PG&E Pri<:e Slmd(alJ) 0.08 0.Q4 0.10 Mojave Amount m:tiv ... 0 0 300,000 • Mojave Pri<:e $fmd(stmxl) 1.00 0.50 1.50 Capital Cost S 1,500,000 150,000 4,000,000 Back-up Cost Slyear 1,000,000 40,000 4,OCO,OOO En.viromncotal Benefit Slmcf 1.91 0 5.00 Pe:formance %tlIrge! , 100% SO"/. 150% mna bM:stmeDI: Cost $ 1,000,0000 300,000 6,000,000 • Opeming Cost $fmcf .73 0.52 1.04 Decline Rate o/oIyCll' 5% 2% 10% • Administrative Cost Slyear 200,000 75,000 350,000 • fERC Cost Slyear 60,000 10,000 100,000 • CPUC Cost $!year 300,000 100,000 400,000 • Seasonal wriables ,.t.kb change over !he evaluation periOO A-<> " .. -.' APPENDlXB Details from DetemiDistic Study The purpose of the dettrrninistic phase is to develop a value model which iinks the uncertainties to an 0llIC0me SO that the altemaives can be ccmparcd and to reduce the number of variables and alternatives in orda to 10cus alli:ntion on onlv those volriables and alternatives that affeet the decision being mode. Two conclusions were dro\.n from the dc:tmninistic phase: (1) The DSM and drilling decisions may be postponed because they do not affect !he critical interstate gas and ttansportation deci.si0lL and (2) The lDJ!l1ber of variables may be redu=l to three. SENSITIVfIY ANALYSIS The "nominal plan" is the optimal policy (the one v.ith the lowee: TRe) ",hen all variables are set to their nominal or base values. In tl-.is analysis the nominal plan ...ncn using me as the aiterion is to purchase domesti<: gas, ~ it on Mojave pipeline, participate in California drilling, and do level 2 DSM The nominal plan for rate impact does not include any DSM The sensitivity of each uncertain variable "as tested by running !he model v.im the target variable at its blllh and low values while holding an O!her variables at their nominal value. The variable is deemed sensitive if!he optimal poll.: changed from the nominaJ plan ",ne" the target variable is set at its high or low exIreme. The pbonomaloo of se!lSitive versus noo-sensitive variables is illusttllled in the IDmado diagrams in labeled Figures ~ 1 and B-2 The nominal plan is designated by the vonieal line in the middle of the diagram. The effect on the 'IRe or rate impact is shown by the length of the bar neoct to each variable. However, only the shaded bars indicate a change in policy caused by changing the target variable to its high or low value. The rainbow diagrams presented in Figure B-3 throogb B-9 illusttate "here policy changes occur and .. hat the policy chan..,. are across the full r.mge of values. Finally, the relatiombip betw= the interstate gas a ... d transportation decision and the O!her two decisiCllS is shawn. in Figures B-10 and B-1I. The illustrati"", prove that no mailer which .. 't<ma!ive is cbooen in the initial decisiOll, the decisirn en DSM and drilling remain.< the same. This relationship allowed the DSM and drilling decisirn to be postponed until the next gas lRP. B-1 sw .. _ -­.......... ........ ~COIt: ~1I'IrIIJICW1:w PG&f~r.a .... ~ ...... fl F"MD W,olf ....... --­...... ---­ FERC_ ---CPIJO_ .......... c..: _ ...... PG&E PGU-._ PGT __ -­ PI'Io.,_.,.r~~ .... PGT __ Value Sensitivity Diagram ~i4kL:_,_, t:::s '- I -,oaxm:o -""""'00'0 B-2 • -T';' _ ~forSWGas ITI1'Iatian Rn _ Bocl<-l.I!> Coo! Oiscat.Int fOr ~ GM: EP T tens Rata Gau.. "_~Co« MQj ..... TI"IIle RD Iwount Storage em Mo!ate ~7rwM~..ate EP LayaIf Villus Produdior! 0ec::Iina Rate Orilng li"NftImef'lt Ccsl _~gC'" !'ERC Coo! _~V_ OSM~an~ CPUCCoo< '-Io~ StDIage Cost Amo\Rlt Sl'orage or. P""'AE: FG&E Storage COld Value Sensitivity Diagram SAR < f-~ , < -' -. ----- -3.8 -3.:2 FIglft B-2 B-3 "·161 SOOOOO j ·162000000 j I ·162SOOOOO 1 -163000000 " ~ ·163SOOOOO 1 -154000000 1 -1645OOOOQ 1 ·165OOCOOO j Canadian Gas Price (% eli"""",,!:rom Southwest price) .. N o'Cl.ina.l .165500000, 'IL. ____________ ---\: ______ -/ 1 ~ PC! -~~l~------~----------~~~w~j~.v~e~--~------~------------ ·1537SOOOO , , -165OOOO1lO ~ ::::1 .16S750000 I ·170000000 1 <J ~ 1112!OOOO..I I!' -172500000 j -l737SOOOO 1 ·175000000 1 -1i6250000 -177SOODOO -1717SOOOO -18OOCK1OOO ·C12 -':U5 -0.1 ·(t05 % Figure B-3 c Back-up Cost (%.\'Oar) PC! 0"05 0"1 0.15 5COOOC 1000000 10,10000 CJOOOOO 2500000 30l'MOO 3500000 oM)OOOQO $)'ear figure e-4 ~5 -'--~ -'61<JOaXJO -16"'iSI:X:.DlJ l~el -18BC1XDX1 ~ I , ' Q.S 0.6 0.7 E5C3ia1iOl1 Figure .B-3 DSM Performance o.a (~d~l Level 2 rNOtDinal oj. ~ Figu:r~ 8-6 B4 1.1 1.2 ,3 Production Operating Cost (S,MCF) _'85<100000~.~ ::=j -16eOOOOOO j '- ~-1862000001' ~~ -~t86400000 ~ \ -llJe11OOOOO j ~~DOQ't dull i -1ll8800000 I -187000000 Nominal Dr!ll i -167200000 I i 1---,--I ' 0.5 -1e5200000 -11lS4GOOOO j -Wi6000001 -165800000 1 -166000000 ~-188200000 J -1611l4OOOOo -1~ -1 "68OOOOc -1 !17000000 -187200000 0.5 Drill 0.7 0.8 0_9 SJI,ICF Figun B-7 Drilling Investment [ToaJ S) 1.1 1.2 1.3 ~ ~----.------(NOminal Don't drill 1000000 5000000 Gas Demand (annual escalation rate) -0.02 -0.015 -001 -0005 o 0.005 0.01 0015 0.02 % Fig'J.l:i!. B-9 8-7 '!J'..,.c _______ "" __ ~~M--r,.rry ......... "~'"'"::i~~c=O.--------~~ -~\~::""~~i.. _____ :'} ,-'''''''-___ ''''''''''_-'r LJ - 1", r;:a;aa. j I.Eo'E. ~ (J ooc:... II ~ ~i~i (J \ """ ~ ~ (j '" --[] F\gIft B-IO ~ i\ \1\ a._'"ASQ I \ la NO~ _____ --;,-, ~ 1'~"'';:'''''~'-----< ~ -r~~~~'''''''''''''_",~",,~~ ____ -<~ ------r"'1 1>-9 ·." APPENDIXC Detlils from Probabilistic Study The purpose of the probabilistic phase is te close!)' examine the boundaries of the uncertain variables, interview experts to = the p-obaDility distribution curve for each critical variable, and run the model using the probabilistic data It was det<nnined that the variables frcm the detmninistic phase CNId be divided into t-"O c:areg<ries: (I) true variables 'Mlose values could not be """"" prior to decision making time and (2) value; that could be negOOared, and th=fore, known at the tim.: a decision is made. For example, the oost of back-up service for the Mojave alternative "as determined to be • ~e va\ue rather 1ha! a true v:niaOle. The variable was set to its low value for ,he probabilistic analysis, 00t it may be inCI1.'3SCd incrementally to determine Palo Alto's breaIa:ven poim with regard til this = On the oIher hand. the discount fcr Canadian gas is clearly • true variable with many possible future Y.!L'es. and further analysis "as done. B.l I'robDIity ~ Probability a1COding is a formal process for quantifY expert~. The cxpat's l.now!edge is t:3p!Ured through • structured inl=iew designed to ensure the consideration of an rekvant infmnatiort The process is designed to minimize ambiguity and to cootrol biases which are inberem in any humans ju<lgm1ents. The procedure utilizes. vmiety of tools and techniques tc simplify the numerical assessment of uncer1ainty. The result and product of the systemaIic procedllre i; a probability distribution which expresses the e>qJC:t's opinioo of the relative likeliJ-.ood of possible OUII.-omes. The finaI step in the eocoding process is to ~ the oontirruous distriburioos for use in the IRP model. B.2 Restits Or tbe Process A gas industry and economics expert was interviewed about the price of Canadian gas supplies relative to Soolhw<sl U.S. gas using the systematic proced1= for assessing probability data In addition, t'Ml publish<d price forecasts of Canadillll and ~ U.S. gas w= obtained: the Cambridge Energy Rfsearch Associotioo's forecast and • MoGmw Hill l'orecIt;t. The probability aiSIrJ bution that resulted D:om interviewing the c:>qlCrt was WiDpIIled against the publish<d forecasts. According to our e>qJC:t's opiPjon, there is a 40"1. probability that the 0CbJal relative price of Canadian gas "ill be lower and a 60"10 that the price will be higher th:m the McGraw Hill fo!=sted value. This forecast was done for PG&E at the time they were justif}ing the pipeline expansicn to Canada. The CERA forecasled values, bo"~er. feU within the 90% confidence range of oor e>qJC:t's distribution. The distribution is sncmn in Figure C-I \Wich is a plot: of probability VC'SUS IRe. CI • , . -"-' r • '" The results of the studies and the L"Iterview were input to the modeL and the alternatives were "valuated again Two more ,1Irlab1es were fO'J:Jd 10 be se!1Sitive for the lr'..uspoi tlIioo decision, the price of Southwest gas and the market price of cspacit)' 00 the EI Paso pipeline. The EI Paso rare and the price of So\;thwe;t gas required additiooal pubabilistio analys!." so the same industry expcn "as D..,n:;ulted to arrive at a probabiiity distribution for the possible values. The resulting distributions are sho"'n in Fig.r.:s C-2 and C-3. The cumuJative probability distributions develcp.'d during the expcn intcrviens were then discreri2r:d or approximated by a few represanative values and associlI!ed probabilities. In this case, !he coo:inuous distributions ,"'re approximated by tree value and associB!ed jrOOability pairs. The estimates used ,,= !he cumulative probabilities of O. 085, 0.5, and 0,915 "i:h likelihoods 0(25%, 50%, and 25% ~"cI)'. These points ore indicated or! Figures e -I IlIrough e-3. R3 Risk Analysis In order to ",a1ll31e lhe risk the utility faces, a cunrulative probability cfislnbution of!he IRe was dc\Ielcp.'d fur each n-.msportatior! option. The analysis siJovI,~ the expected value: fur each aItemalive's IRe and all the possible valw:s of IRC for each all<mItive. The results can be seen in Figure C4. . SC\'mII things C3tI be delermincd from this analysis: (I) ~~ave and PGT have expected 'IRe values ~ SI5,OOO Ies., than the otber alternative;, (2) W.ojave has a slightly higher expected value fur IRe, and (3) at the 9O'!'o confidence level M!Jjave 's me is ~ 10 be lower than PGTs. Nominal values for the costs still being negotiated were used to develop this risk profile, Figures C-S and C~ srow the armuaI costs for Mojave and PGT by coot categ<Xy. figure C-7 in an annual TRC <:ornpIllisoD between the two aI\ernatives. The data "'as obtained by using the expected value fur each probabilistic variable .m the nominal ,a1ue fur all O!hc:r variables. The cumulative probabilities are lost in this anal}sis. The graphs were goneraIed solely fur the putpose of comparing !he relative si2e of the various coot categories for each alremative and to show the relative value of coot on an anrrual basis fur the top two ~ C-2 --...,. .~------ . '. .. '-~ " ," . I , -' \ I , I " I w 1N1UtVII,Wtll KLlria D~lluy r tl~ pnO~Ofllr'V fHCOOlN(J "lRM B1i1t.'H;r 1((111 HOc/hflar nAIf. HJIll/')) __ 0,0 O~ 0,1 un 0,2 II , 03 1I fJ • 0,1 O.!,i 0,5 t! 4 08 fI :1 (J_T II ;, 08 41 1 o. 1,0 II" I" i I' i i" I :/ Iii II; li I' HI: I 1Eili1Lili1JJ:iJ , II i Ii 1111 : II Iii I: IHIll (l.{1 0.7 O.H II.!' I,ll 1.1 1.2 I.) 1,5 1,4 Irlj!;lIrc l;-I V,AItIAlfI I 15-Yi6'ur 8veragf~ hltenHitllHllll b(lnl~r Ilrlc" uf (;.lll/l,llali gaH bnuml for Caltflwnia :t '"ellltl_vA to I'enulllll prke!J. (1UIIll-JpJlt!r) I;I)MMI_UI -'-'/ ',{.. 2 c it 0 !E 8 <.> :0 .. ,. ~ .. ~ " L I~ ~ 0- ~ -c w i< c w v ! <; • c g " o ci ~ l1li rI7 ~ ~ ~ ~ ~ ,~ ; ~ c .. = ". Q " 0 ~ " o ~ ~ .. e .. 0 '" i'" '0 " - ~ §: ~ • " ~ ~ ~ ~": "", '=c ~ . I ~ ~ '" -, '~c:3'c -c':=.:c',', ~,,~: '";.: .. :~:=c ~cC: O';=f ;:',c c N , ;c~=' '" on ~ , .. 0; e ~ " = ':: e C-4 -'-' - 2 I ~ • • I 0 " • -t ~ • • ~ '" " 1 , • ~ = • • • .2 • • .. • ~'" o-w ~ • • 0 • ~ = ~ • --• • > ~ , • -• 0; • " .-~ >< ;; ;; ~ • " i > ~ '" N ., • .. '" ~ '" '" .. .. .,; ,.: C C C c <' ,.: .. ~ ~ .. " 0 • 0 , ~ ~, ~ -0 <> ~ '" Z -w ~ c > '" -~ W ;; :;: .. " " Ii .. I~ '" '" 0 < .. e c. "-- ~ • • • ~ " • 'ii • • • 0 § C ~ 0 " ~ {; ;; E ;: '" S: ~ "'"' ~ c ~ g ~ " '" ~ • "5 " g ~ " ~ " " • > • 0 • • , • , " , • • • ~ :; ;; 0 • " " • • '" , ~ E • ~ • 0 , " > • li • , ~. • , • , " • , i-i: ~ • :! ~ ~ :;; ::w c " = ! ~ ~ " > ::; c " " " ., ~ " .. ~ .. ~ .. = ~ " " " " c '::-3 --~-. , .. ..-1t e.,. "-. , I I ... ' ...... 1 -,1." 1 J., .. ' , RISK ASSESSMENT OF ALTERNATIVES Figure C·4 1 r ---i--rr --·----------I--_>-~ O 9 I I i I -i---;:"--··-l • 1 i ' , --, .' . I i --------,! I-~----·--- 0.8 ~ 07 ..J . iii ~ 06 o . ~ n a. J, ~ 0.51 ~ 0.4' :::> ::; ~ 0.3 o 0.2 0.1 o ,! : /: ! ., . '. , " t , " i ; , ., , ., , ., , ., ,/ , , , ./ J' .. .: " J / /~. --'--:....L·T -·--- 100000000 J _--+~:-----::~"">/. D~~~~~: ! J~ ,,~' ~' ! ,./ -:1 1 .'. /: I'! ! ,. ;, ! / : t' ! i I /!' , , : ! i I I _11_ .i .. 150000000 --·1 200000000 TRC T--Ti 250000000 300000000 DECTRAN: PGT DECTRAN: El._PASO DECTRAN: CS • ~ f 20,000,000 ! 18,000,000 .:. I 16,000,000 t 14,000,000 + 12,000,000 (i -- 10,000,000 8,000,000 6,000,000 I 4,000,000 ! : I I j . I ' , j f ' ~!; , 2.000'00; JIIIIIII.III.III 20,000,000 ~S,OOO,OOO 16,000,000 14,000,000 12,000,000 10,000,000 8,000,000 6,000,000 T T ! t 2 3 4 5 6 7 B 9 101112131415 Hi YEAR Ftgute CoS PGT IPG&E COSTS -_ -=i -:..-=:: _ -.J -......".; --....; ---., ----i _-~ -:--~~- 2 3 4 5 6 7 8 9 10 11 '2 13 14 15 YEAR C-7 ---.~--.--.--- =:J Capital cost 3 Admin • FERC • Gas .cost C Interstate transport • Intr.ilstate transport. • Storage :::J CPUC • FERC • Gas cart = lnter3tare transport • Intrastate Transport ----.~p----- Mojave VI PG&E Total Resource Cast 20.000.000 T 18.000.000 + 16.000.000 i 1'.000.000 I 12.000.000 ! , • ?G&" 10,000.000 8,000.000 = Mojave 6,000.000 <.000.000 2.000.000 a 2 :3 4-5 e 7 a 9 '0 11 12 13 14 15 YEAR FIglft C-7 C-g