HomeMy WebLinkAbout0137.093-'.
:.,~ ". -'~ . ~ .:"'"
".' ..... ;...., ..
, ..
~ ~
J~nuary 21, 1993
THE HONORABLE CITY COL~CIL
Palo Alto, California
Electric Utility 1992 In~egrated Resource ~
Members of the Council:
The Utilities Advisory Commi~sion (UAC) is forwarding the 1992
Electric Integrated Resource Plan (IRP), prepared by staff, tc
the Council fnr their information. No Council action is required
at this ti~e. At previous Council meetings, staff has requested
council approval of some of the r€commendations o~tlined in the
1992 Electric I~P. This report i~ a compilation of the planning
:metJ":,od., asswo.ptions and res\Jlts of the IRP.
Piscussion
The National Enerqy strategy, passed in the fall of 1992,
requires that Western Area Power Administration {WAPA) customers
adopt inteqrated resource planning (IRP}. WAPA, the City's major
electric supplier, is requiring their customers to use an IRP
process that will simultaneously evaluate energy efficiency
proq~ams and supply alternatives in developing thei~' resource
plans.
With an eye on the future, Palo Alto responded in a proactive
manner. In ea!:"ly 1991, staff began developing the City's firs.t
IRP. The goals 'Were to minimize resour'ce cost to all customers,
air pollution and upward impact on rates. These goals wer~, at
times, in direct competition; therefore, staff's approach was to
carefully balance these competinq interests, givi~g due
consideration to uncertainty of future events, in order to
establish a plan that is robust, environmentally sound and
fiscally responsible.
The result of that effort is the 1992 Electric IRF, which is a
planning document that serves as a 5hort-te~ action plan and a
long-term guide to the electric utility'S resource acquisition.
It is the CUlmination of a year's effort that incorporated the
expertise of the utilities Advisory Commission ana input from
industry experts and the pool ic. The short-and long-term.
recommendations of the 1992 Electric IRP are:
ClOt: 13 7: "
,
Short-term Action plan
Participate in the Seattle City Light Energy Exchanqe
Do not participatf'! in the HCPA Combustion TUrbine Project 2
Implement pilot DSM programs with OSM lev.el 2 (30 MW) fiS the
long-term target
Long-term flexible plan
Participate in studies of Calaveras {French • Ram5~y) hydro
enhancemen~s
Participate in feasibility studies of annual baseloaQ qas
project and summer-only baselcad qas project
Pursue 75 percent of OSM achievable pot~ntial (30 MW)
Investiqate local generation and co-qeneration options
A few of these recommendations have previously been brought to
Council's attention for approval, and some have been altered in
light at new information. FUrther recommendations will ~e
brought to the council at the appropriate time.
The IRP will be updated every two years to respond to future
changes and regulatory mandates. This should help Palo Alto
achieve its ~i56ion of providing q~~lity electric service to our
customers in an economic~ efficient and environmentally sound
.anner.
Respectfully 5ubmitted,
c::rt'c,-+t"b,,~,I-I---
TOM HABASHI
Actinq Enqineering Manager
Resourc~ Planninq
,,-~ A_ He 7
RICHARD L. YDUJIG
Director of Utilities
Attachmentt Integrated Resource Plan
CHlt:U7:"
-
,:
, '" .. . ",
!If
,;
~
~
,;
. '" IW' ,;
.;
'" ...
'" If ;
'" I!r
'" Ill'
. .,.
I(
'" .::
'" II.
'" l ..
It ..
II, ..
' ..
~
~
':.
~
~
~
"
CITY OF PALO ALTO
ELECTRIC UTILITY
1992
IITEfJRA-TED-RESOUBCE PLAN
SOCIETA.t
COST
EHICIEWCV
~ES<)U1;:CES
~ATES SWO~TT£~
ACTIOWS
.. jl .•
· .••....•.. c •..•.....
1ii
"," --
i
I
j
i -...
. ,.
TABLE OF CONTENTS
Table of contents
List of figrJres
LIst of tables
List of abbreviations
EXECUTIVE SUMMARY
ClIapter 1 OVERVIEW
U
12
1. 2.1
1.2.2
1.3
1.3.1
1.3.1.1
1.3.1.2
1.3.1.3
1.3.1.4
U~
1.3.2.1
1.3.2.2
13.2.3
1.3.2.4
1.3.2S
13.2.6
1.4
Ifutory of the Palo Alto Ele<tri, Utility
Iptegrated resource planning
Differences from traditional approach
O\7elView of Palo Allo's IRP and decision QnaO'JiS philosophy
Palo Alio'. 1992 IRP
J\1"otil4lli,;n
Wutern alloco1ion.
NCPA COlflbusriun Turbine Projecl 1
Preparation for foil-scale DSM" programs
Seattle City Light Exch4.nge Agreemenl
Process
Evaluation(screening of supply resources
EvaJualio1J/screening of dem~;lnd resources
Problem formulation
Deurmiltlsric llNJlysis
Uncertainty analysis
Results and recommendations
Overview of the document
a.apter 2 LOAD FORECAST
2.1
22
2.3
H
Forecasting methodologies
Historical etectricity ~ and forecasting methods used In Pale Alto
1992 load forcc=
Future forecasting methods and efforts
Chlpl .. J EXISTING DEMAND-AND SUPPLY-SIDE RESOURCES
3.1
3.1 I
312
3.2
3.2.1
Demand-side resources
PARTNERS
Load management
SuppIy.side resources
Western Area Power AdminiJ/rolion Contract
Lily of Palo Allo 19911RP
IlL
lV
V
'0'11
1
8
9
9
11
14
16
16
17
18
20
II
21
21
21
22
. 22
3,2.2 Cameras hydroeJec1l1c project
3.2.3 Pacific GQ~ & Elcctric off-peak ill/ergy
3,2.4 Was.S,ington Water &-Power contra,:t
3.2:S Pacific Gas & Electric partial requiremem
3.2,6 CaTifornia-Oreg(.·n Trommissiol'! Project
3 J Load and resource balance
Chapter. SUPPLY & DSM R[SOURC~ SCREENING
4.1 Supply-side resources
4.1 I Supply-side reSQurce screening "pH' IRP
4.2 Demand-side resources
4.2.1 PreTiminar:y ~;creening o/NCPA F.FP proposals
4.2.2 Second screening o/NCPA REP proposals
4.2 3 De'Velopment oj in-house programs
4,24 J.fodelling in-holA,Ie programs
4.2,5 .Analysis olin-house pmg'Oms
Chapter 5 ENVIRONMENTAL EFFECTS
S 1
S.2
s.J
.Reasons to consider environmental effects in planning
Valuing environmental emmalities
Extemalit/ val"es used by Palo Alto
Chapter 6 RESOURCE INTEGRATION AND RESULTS
6.1
6.1.1
6.2
6.2.1
6.2.2
62.3
6.3
6.3.1
6.3.2
6.4
6.S
6.6
Detenninistic Analysis
Estahlishing the size of supply ~sourceJ
Detennining the five alternate plans
Analyzing all combillQtions of supply resources arot} DSM levels
Ranlcing u.e results of lhe all-cQ",hinalion analysiS
Heuristic analysiS to select jive plam
Uncertainty analysis
Identifying die uncertain variables
Determining !he sensitive uncertain va:riabJes
Assessing probabilities of unceita.in variables
Con..s:tructing the decision tree
ResuJts and recommendation:;
Appeudi. A
ApJKDdiI B
Appendh C
lessons learned and ideas for improving the 1994 IRP process
Description of DSManager model
Description of Multi-objective lntegrated Decision Analysis System
(MIDAS) :nodet
AppeDdix D Del.a!led results of deterministic analysis to determine alternate plans
Of)' ofPakJ Aria 1991IRP
22
22
23
23
lJ
24
27
27
27
29
30
)0
31
33
33
36
36
38
38
40
40
40
42
42
43
45
50
50
S3
63
65
1;6
<I
~
i
•
•
• I 1
L2
1.3
L4
U
1.6
2.1
2.2
2.3
3.1
3.2
6.1
6.2
6.]
6.4
6.5
6.6
6.7
6.1
6.9
6.10
6.11
6.12
6J]
6.14
6JS
LIST OF FIGURES
The decision analysis process 4
lntegrated reso urce planning process 6
influence dilgr&r.l identifyi.ng ur.certaintJes & value crirerie. 8
Decision tree ior probabilistic pha5e I 1
CummuJative probability cfistribution of alterna1e plans using SA.l(. and societal ccst 12
Expected and extreme vllues of aItema:e plans using SAR and societal c.ost 13
20 year en.ergy forecast by ~tDr
20 yeJIl demand fo re<1ISt
20 year eder gy forecast
Load and resourcc balance a Existing !'esources
Load and resource bsllllce ~ Pisn resources
Decision tree to establish siu of supply resources
Decision tree to determine alternate plans
Influeo~ diagram identifying uncertainties and value criteria
Decision tree to determine sensitive uncertainties
Sensiti\lity to for~otf: load
Sensitivity to hydro level
Sensitivi!y to WAPA energy allocation in 2004
Sensitivity to DSM .,..,etration
Sensitivity to WAPA capacity aHacarion in 20Q.1
Sensitivity '0 natw"al gas price
Sensitivity to DSM impact
Sensitivity to annual baseIoad gas cost and fuel cell avaiJabillty
Decision tree for probabilistic analysis
Cumulative probability distribution of societal cost and SAR
Expected and e-xtreme values (If societal ~st and SAR
Overview of the analytic framework in DSManager
17
19
19
25
26
41
43
51
54
55
55
57
58
S~
6{J
61
62
66
67
68
B-1
CIt)' 0/ Palo Allo !991lRl' iii
-..
. ,. -,
•
LIST OF TABLES
U ShorliS'!~ supply and DSM resowce data 7
1.2 Resource capacities of five different pians 9
1.3 Uncertain variables and sensitive uncerta:nries ;0
1.4 Recommendations 14
4.1 NCPA sbonI;st 28
4.2 Palo Aho suppIY-!'e"..ourte shortlist 29
4.3 NCPA DSM proposal shonIist 31
4.4 In-bouse DSM programs 32
4.5 Cost-beoefit of individWil Palo Al to programs 34
4.6 Palo Alto DSM levels 35
5.1 Impact &5SO(:iated with etettrictfy generation from different fuel types 37
;.2 Externalities developed by CEC 39
6.1 Upper and lower boundaries of supply resources 42
6.2 End.points occuring Wider T RC and SAR ran king 44
6.3 Suppiy resource frequency table 45
6.4 Dem~d resource frequency table 45
6.5 Step I in deterministic plan seiectiOD 46
6.6 Step 2 in deterministic plan selection 46
6.7 Step 3 in detemUru stl<: plan selection 47
6.8 Step 4 in deterministic plan sel ~O!l 48
6.9 Step S in determi.nistic plan selection 48
6.10 Step 6 in dete~istic plan selection 49
6.11 Step 7 in deterministic plan selection 50
6.12 Uncertain variables .end sensitive uncertainties 63
6.1l Probabilities of snortliS;:ed uncertain variables 64
614 Five diffe..rent plans used for ana1ysis 65
615 Recommendations 70
Ciry of Palo A.lro 1992 !RP IV
.. . -
..
.. .. ..
' . •
HVAC
IRP
MIDAS
City of PDW AIM /992 lRP
LIST OF ABBREVIATIONS
v
Cil), of Pale Alto 1991 lRP vi
o·
.1
•
EXECUTIVE SUM'>I..I.RY
Electric utilities throughout the United S!ates are ernpatklng on a new era.. Across the nation.
res:ulatory agenctes, pelitica!. z..ction groups. citizer.s lIJtd industries ate demanding that their
utilities, investor owned or COn5U-'l1er o'Wlled alike, not only build resources to meet electricity
d~mand but aJ.so be proacti'li't in Shaplf\.8 that demand in Nder to minimize oost and
environmental damages to the community and :so<:iety at large
The United States Congress is IIlcreasingly motivated to pass laws that are protective of the
nation's natur&l1eso"U1ces and sup?Ortive of .actions that are sensitive to the env:ronment. In fact,
the National Ener~ Strategy legislation passed in th.e fall of 1992 requires that Western Area
Power AdministratioD (W.b..PA) customers adopt Integrated Resource Planning (IRP) In
Cahfomi., the Ca.lIfornia Energy Commission is scrurinizi:1g the rewurce pians of utilities to
LI1SUTe that resources are built only when they are needed WAPA. the City':; major electric
$upplier~ is in the process of requiring their customers to use an IRP process that wiH
simultaneously evaluate energy efficiency programs and supply s1te~atives in developing their
resource plans.
With an eye on the future., Palo Alto is responding, as it alw.ays has, in • proactive m.mner In
early 1991, staff began developing the City'S first [Rp, Our goals were 10 minimize resource
cost to EJI customers, minimize air polh .. i.ion and tr.inimiu upward impact on rates_ These goals
were, at times, in direct competition; therefore, staff's approatb was to ~arefully balance these
competirtg interests, giving due consideratiOr'l10 uncerta.tnty of future events. in order to establisn
• plan that is robust. environmentally sound and fiscally r~ponsible.
The resl!11 of that effort is the 1992 Electric IRP. This document serves as a short-tenn action
plan and a. lon8~term suide to the tl~trie utility's r~ur-ce acquisition.. [t is 1he culmination of
a year's effort that incotpCrated the expertise of the Utilities Advisory Commis£iorl and input
from industry ""Peru Bl?d the public. The IRP will be updated every two years to respond to
future changes and regulatory mandates. This should help U5 achieve our mission of providing
quality electric service 10 our custom~r!i in an econo'!1ic, efficient and environmer.taHy ;ound
manner
City 0{ Palo .. filo 199} IRP
~
"
•
I
I , ,
1. OVERVlfW
Th~s chapter provides .. broad description oi the 1992 integrated resol.ln;:.e plan (IRP), including
results and recommendations. Beginning with I brief note on !:he history or the c:le<:mio': utility,
the following S«tion~ compare previous methods TO the IRP. and expl&in \\1J.y Palo Alto was
motivated to develop an fRP io 1992. Also inctuded is a short discussIon on decision analys;is,
presented in the conteKt of an overview of the IRP process. Further detad (In ea.ch phasl'! of the
IRP pro-cess is a.ccessibte i!'l the er.surng chapters, and an outline of the entire document can bC'!
foumj at the end of this chapter,
U lIistory of the Palo Alto [Iett,;c Utility
It was the futuristic: thinking of two Stanford University professors that was larg.!ly
responsible for the "mergence of muni cipally-owned eo] ectric service in Palo Alto at: me TUrn
of tho centwy. Professor Charles "Daddy" Man< and Charles Benjamin Wing led the
movement to operate municipaJly-owned utj[iti~ in the fted.gling town, The)' freit,uendy .
atgued that the CLty could provide electric service at rates significantly below thc!se char-ged
by private ele<..'trlc companie1.. Such arguments not only proved convindng an<f. sen.'ed as
I catalyst to municipalize elt'ctrlCLty, bu! more importan.t1y have proved ttl be tnle today as
weU as then. It is a1~ wort:h noting that one of the foundIng principles of these early
piolleers is that the utilities operations must show a profil.
The idea first became a rea1ity on January 16, 1900, in the fonn of. 12S horsepower
electric steam generator. The initial $12,000 funding for Ihe plant came from the unused:
proceeds of a $40,000 seMlr syst~m bond issue. However, this plant could not meet all
of the to\W'.s demands, so supplement!! power was purcnased at 'W'holesale rates from die
competing private utility, United Gas and Electric which was subse'luenijy absorbed by
PlICific Gas and EI ectri c Company (PG&EJ.
In 1914 the first diesel engine was instalted at the City's powemouse localed at Rinconada
Park. In 1923 the City connected 10 PG&'E's system in order to met1 rising electrical loads
causea by its population growth. As the dem.1lld for electricity continued !o grow In later
years. the diesel engines were used primarily for peaking purposes. By 1930 OIi]Y 5% of
the City'S pewer needs wert generated ~ithin Palo .Alto at the power plant, while 95% of
its total requirements were supplied by PG&::E. rn 1941, hced with the reality that the
dieselmanuIactwers no longer supplied p!lJ1S ror these earlier engines, the City sold them
for sera:p value.
In 19641be City ceased its wholesale power pur(ha:res from PGkE and began pwchasing
its entire ~wer needs from 1he Central Valley PfOj~t (CVP) o-ptra.ted by the federal
gClvemmen.t CVP power is marketed by the Western Area Power Administration
{Western}. In 1983 the City Council etected to participate in a 2-3~l. share of the Nonhem
CalifomiaPower Agency (NCPA) 230 MW hydroelectric project (Calaveras), which began
~tr1li",'If'
Ciry of Pate Alto J 99] lRP
op~rarion in 1990, In August 1984, the ele<:trical requ~rernenls of Palo Al'!:o ex<:eeded 115-
contracted le ... el with Western a.,.c1 the Clty received parti.a1 reqUirements power from PG&E
ror the first time since 1964. From 1985 to 1990, the City m~de! major romr.lLtment to
conser;ation lnd load management The Clty h.as alS(' ap~roved p<lwer p,.!.rchase contrac:s
'With utilities in the Pacific !'lortbwest. A comerstone of the City's resoun;e plan fo. the
present and well into !he future is the rehance on electric: capacity supplied from We:itern
and the Calaveras Hydroe!ectrii; Project.
1,2 IDtqrated Rtsource Plannin& (IRP)
The customers' increasir:g el~ctrici!y Deeds can be met in two ways The (Lrst, p,ppiy
f'eJQurces. is to construct and mainwn gener.ators, tra.'lsmission lines and dlstribution
systems kl adequately and r~liab[y meet the customer's demand, The second, consen'ation
or demmtd-side reSOlln:es. is to encourage and persuade customers to ~ energy more
effidently~ thereby freeing up the savings to meet new demand. In effect,. cot1served
en ergy is considered its own resource.
In the Past. energy planners relied almost exclusively 00 supply !e50UICe:s to meet new
ciemlUld.. Today, research shows that demand-side resources c.m often do the job for less
money. Since standard planning modets oould l10t effecti vety deal with both supply and
demand~side resources, 00 a level play!ng field. a process was developed that could -
Integrated &source PIoi1ning.
1.2.1 Differeocer: from traditional .ppreach
In the past,. supply ie50urces and demand resourl7.S were evaluated in two separate aren3.!] ,
TradltionaJly~ supply resources were evaluated individuaUy using electricity price as the
sole det~rminant for resource selection.. Demand resources were implemented only if the
sum of th e demand resource cost and the revenue from lost sates was less than the
alternative cost, n8r.lcly buying supply resources.
On the other hand. IllP eva.luat~ supply and demand resources simultaneously, using the
criteria of electricity pri<::es, environmental: effects, and the total CQst of energy supply
from the combined customer and utility per5pe\.""'tlve.
The IRP planning precess is a coordinated effort that inspired participation from various
utility divisions, the general public, the Utilities Adviscr)' Commis .. ion (U AC) and outside
energy expens. The tradltionaJ planning process took place 'Within a single division.
withcut th~ checks and balanc.es and the fuil spectnlln of inputs and exp.ectise that the IRP
process encourages.
(h>tn>jr.
Cit)' of Palo Alto 1992 J.RP 2
,-.'
•
•
1.2,2 Overview ollRP and decision anillysi! philosophy
An IRP is both. short·term IlCtion plan ofthingi we will do without need re,., further study
and :t long range conceptual plan of things we If'lticipate doing. If supported by the results
of ruto.re more detailed analysis The long.term plan is therefore fl"'Xlble and responsive to
chlU\g~n.~ informaticn
The ir.tegrall?d part of lRP refers ;0 the prcces.c;: of simultaneou£ly sut-jec'cing supply and
demand resources to rigorous analysis and then measuring thP,: results l,I,ith the sa.me
yardstick~. The reselting mix of both types of resources helps to ensure thas the customer's
demand for etectricity will be met at the leoast CO~1:. For Palo Alto, the chosen yardsticks
were societaJ COst! and electrici Iy ate.
Proper integrated re;;owce planning helps utilities sift through th~ variety of d~rr.and and
supply options and end up with I low-cost. stable energy supply for the long run. An IRP
can guide utility action s to mw mize benefits and minimize burdens for users of electricity
and all those ,ffected by its use. To that en~ an IRP recommends actions that avoid
e:occessive 'JSe of unstable resources and encourages the use of resources that have the least
adverse :affects on the environment.
A:n. in:egral part oJf the IRP de.als with risk consideration. Decision QNJiy.sis (DA) is •
methodology used to incorporate risk ir.to the equation. Figure-1.1 shows. graphical
description of the DA p~ess. DA takes a complex problem and arran6es the individua1
pi eces into a series of well defined steps, diSGrete b u.t not necessarily independent The
analyrica! advantage of DA stems ftom its use of a dtcision tree to represent all these
different steps. A decision tree is & visual tepresentstion of the probiem. (onsisting of a
combination of decision and chance Daees. This formulation belps to CI)[T .. municate the
process used to evaluate. analyze and eliminate alternatives. clearly mapping oul the path
from probtem to solution.
DA provides aD excellent frarde'NOrk for the IRP. because the IRP is • complex problem
requiring input from different divisions of the \ltility and encouraging additional input
fram the public Structuring the problem into. sequence or discrete sta.ges helps each
parti c.ipant in d! e process focus on I partic!.11 at element, depending upon t..i.ei!
responsibilities and frame of reference.
Chapter 6 provides mor~ infonnation regaIding DA techniques and the results we, obtained
usinS !hi. tecluti que.
A g.-eat deal of effort went into determining \Which f.actors wouJd have the greatest
influence on the plms, with an emphasis on resulrs that stand up to the uncertaintie.s of
future and as yet unknown events. In other words, a rombination elf adyanced computer
modelling ar..d I. wmprehensive undtrstanQng of energy supply and demand history
O\'lI!n'iew
Cu)' of Palo A.l.o 199} IRP J
Tbe Dtdsion Analysis Process
Fi8'= 1.1
An explicit q'J<L'ltification of environmental externalities was used to penalize resowces
cauring air pollution, Uld the environmental impacts of each resource were considered in
arriving at the final recommendation Other externalities will be incorporated aco they are
developed
U Palo AIle'. 1991 lRP
1.3, I Motivation
There were four pres.coing reasons to develop an IRP in 1992.
1.3.1,1
L3 1.2
Ow",i.nf
Western .notation.
In 1991 staff submitted joint comments with the Sacramento Municipal Utility Distrid
and the Natural Resources Defense Council to Western Those comments, which
require prepatatioo of an IRP and a commitment 10 improved energy effici.:nc;y as i
requirement for allocation of W.estern resoun;es jn 2005, were incorporated mto the
proposed Western Energy Planning and Management Program (EPAMP), The Palo
Alto IRP meets one of me requirements set forth by WCS!:ern's prop<tSe1i E:.l'AMP to
obtain renewal of lhe City's existing allocation. The remait\ing requirements will be
met as the City follolN'S through with implementation of the recommended Short-Term
Action Plan wInch will b. periodically updated and sobmir.ed 10 Western
NCPA Combustion Turbine Project J
NCPA was in the second pba.se of developing a 49 MW steam injected gas turbine
{STIG). The City, au a participant in the seco~d phase agreement for this project was
eil}' 0/ P,.lD .4110 1991lRF 4
•
•
I
1.3.1.3
U.1.4
•
"
entitled to up to 200/0 of the Wtit's output. approximately 10 MW. The SDG j:L"Oject
was evaluated in the context ofIRP and hence had to compete with o~er supply· and
demand·sklo! resource 'Options that would be available to us as an outcome of NCPA's
.JI·sou;ce Request for Proposals (UP}. The STIO tilird phase agreement was to be
signed in August 1992 and the CIty needed to mak~ adecislon regardmg partiCipatlGT1
by July 1992
Prep. ration For rull·,suJe DSM prorram:l
The 1992 DSM report investigated the technical and achievable potentiltl of DSM.
It was O~vlOU5 from the res01lts of the report that there was I-promising opportu1\if)"
to include additional DSM as pan 'Of the resource plan. While the DSM report
outlined the budget needs far the 1992-94 fiscal years, it had nct specified the
individual programs for implementation on a pilot or full-scale basis_ The completion
of the IR.P gives. the DSM imptementers adequate time to &llOC&te their budget in
order to design pdot programs and other DSM programs. This should belp position
the City for fun scale implementatioc to aclUeve the recommende.d Iong·tenn targets
specified in the IRP.
Seattle City UJht (SCL) [«bAD,_ Acreem •••
During the initial stages of the IRP preparttion. NCPA began negotiating the SCL
«change contract Negotiations were OD i. fast track, and required I. de<:isioll on the
level of our participation by August 1992.
1.3.2 Pn>eOSJ
Pigwe 1.2 slKIws I. flowchart describing the IR.P process staff foHowed
The-following headings provide I. one-line description of the steps taken in camp] eting
the IRP. This is followed by • brief de$Cription of each step Subsequent chapters deal
with each of these steps in more detall.
Owm~ ...
Ev.aluation .... screening of supply resources
.Evaluation/screening of demand resources
Problem formulation
Deterministic analysis to develop plans and s..'lorthst uncerta.i.nties
Uncertainty analysis· probabilistic phase
Findings and recommendations.
Cily ofP~ Alia 1991 IRP s
>
J
" .
Integrated Resource Planning Process
DSM Report NCl' A DSM
RFP
NCPASupply
RFP
March 92
er Supply I
o()Q-si~ s~!it>n
oCo-gen::T!ltiC[I March 92 ~h 92
Ov.c-1"";of'W
DSM:mager
EPRl tofl:wue 10
1JCIt'II"'.J DSM me.a.sureII
SctnariolStrategy
Development
·Scena.rios ;.f future strategies
·C<>mbination(s) of resour=
!oevaluate
MIDAS & DPL
DPL is 10 focus analysis throogh influence diagram coostructioo 8< ! deQsioo _ pr_tion MIDAS, an EPRI prooU<1, is 10 model productiQll
~ rew:nue requ:irernems and murticipeJ filWlces. Decision tree-front
end and qulck execution times allow for rapid turnaround Probabil istic
analysis can also be perionned wing MIDAS.
figure I 2
City f)j PalO' Alrq J991lRP
. ;"
•
6
_ .. .. .. .. .. .. .. .. ..
•
•
•
•
•
•
•
•
•
•
1.32.1
Future supply resources were examined using infunnaric[l from SePA's aU·source
Rr-P and in-house studies of !>Otential alternatives_ .It.. spreadsbeet model was used to
screen this list of alternatives A total of seven suppiy T!:SOurces were f~und to be
suitzhle for further de~led evaluation. One was a share "r NCPA Combust.on
Turbine Proj~ No.2 {STIG), two were contracts for natural gas fueled g~ntrators •
one was the Set Exchange Agreement, two were en.f:.ancements to the existing
Calaveras hydrotlectri.c project,. and one was the evolvins fuel cell technology. Tibie
] 1 provides a description of the characteristics of these resources .
Table U SHORTLISTED SUPPLY'" DSM REliOURCE DATA
1 Resourc:e Year Capscity EDo!:I'V Capanry CC'st
Da-]in~ Factor Find Variable
MIl' MWH % $i7cw-yr $Imwh
STIG 1995 5 37.)0 85 112 26
Sel 1993 II 9600 (a,b) N/A (e)
B_gas 1998 6 47300 90 139 13
Summer gas 1<;98 6 26300 100(d) 120 14
lWnsey 2002 J 7000 (e) 276 3
Frencb 2000 3 9600 (f) 258 3
Fuel cell 2000 2 16600 95 295 22
DSM levell 1994 20 84000 48 4.77 10)
DSM lovel2 1995(8) 30 139600 53 4JJ 12.54
DSM level3 1996{h) 33 170800 59 4.73 1343
.) Debverics from SCI. -Ju.ne throagh October. CF = 50%
b) Dc:iivcriel from NCPA -No,'ember through April, CF "" SO% to 70%
e.) Vwble cosI it 'Ystet.l inc:remec.W eUCf8Y I:GSt
d) Avaibble 100% -May Ihroush October
e) VL<in from 13% to 66% o Varia from IJ'Y, t!I 86%
g) S!IiCe DSM klld 2 ill. "supenet" or level I. oD-IU-.e date refers w program! in level 2 lMt 4fli bOt pR:Mcl
i.e [eve! I
b) Sinu DSM level.): iI • ·supenet" of Level 2, oll-Une date refers to progrllmS in level 3 that are oot present
in level 2
Ov.,..,,:'rot/
Clly ofPIl/oAltc J99} 1RP 7
1.3.22
1.3.2.3
Oven.it''W
Western
c:apacity
oInocation
Evall1'ati(ln/StrteDin: of Demand Rescuro:es
The 1992 DSM reper. W3.S I detailed S't"Jdy of the po!e~tial DSM (\opportunities
",,-itbin the City. The cost..-effectiv~ measures were de"eloped into twdvc
programs. These :welve programs were then groured into three cumulative
levels. Levell targeted 4~% afthe acruevablepoten'iaJ.levet 2 indildecllevel
1 programs and targeted 75% of the liI.chievable pote ",nal and level 3 included
level 2 programs and targeted 9{)% of tile achievable potentia! Table 1.1
provide!:i a description of !he chara.c:teristlc-s of these resources.
Problem Formulatio!1
Fonnulating the problem COr.iS~sts of stating the problem or decision to be
made, identifying the uncertainties that could affect the deCL5:ion and stating the
criteria used to evaluate the results of the analysis,
rNFLUENCE DIAGRAM IDENTIFYlNG
UNCERTAINTIES & VALUE CRITERIA
Figure] .3
City afP,,/oAl!o J99J lRP 8
•
1.3.2.4
Table 1.2
PIaR A
PI .. B
PIa. C
Plan D
Plu E
1.3.2.5
(h_nri,,,,,
•
The problem is stated as follows ~ "Decide on SnG participation (0 or :S .MW) in the
context of a 2-year .:.chen plan together wirh a 20·year p!/Ul from the lRP that \\-"'ill
COris] der demand III d supply al tematives.·
The uncertainties that were identified are natural gas prieto load forec~t. bydro kvel.
Western energy allocation in 1005, extent of DSM participation, Western upacity
aHocation in 2005, DSM impact and rescurcl: a .... .ailahi 1ity and cost,
The-criteria used to evaluate the results are societal cost (SC) and system average rate
(SAR)
The influence di!1g1&m shown in Figure 1.3 identifies the uncertainties and value
criteria that \Vere used in the analysis
The screening process left us 'W:ith severt supply-side resources and thr~ DSM le .... els
'Which contained '2 prOcgrams A combination of strit;t analysis and heuristic
judgement was used to combi.ne these resources iilto five alternative plans, shown in
Tabl. 1.2
RESOURCE CAPACITIES OF FIVE DIFFEREi'iT PLANS
(All units in MW)
STIG SCL 8GAS SGAS FRENCH RAMSEY FCE_~DSMI
~ II 6 6 0 0 o 30
0 11 6 0 0 0 2 J3
0 9 6 6 0 3 0 30
0 9 6 6 1 0 0 30
S 9 6 6 0 0 0 20
Vnumiat}' Ane1YDI
This step detennined which W1(;ertainties, identified in the problem formulation pha'5e,
were sen5itive and ought to be subjected 10 detailed probabilistic analysis. This was
done by observing if the prefetred plan changed LD response to setting the uncertain
vanables at their high and low values. Since the preferred phm remained the same
in most cases, we proceeded to shonlist the uncertainties based on their ir.cremeiltal
ell'y uf Palo AI.o 1992 IliP 9
• '-
impilcts on SC and SAR. Table 1.3 lists the uncertainties that were identifled and th~
four deem~;:I sensitive.
Tabl. U UNCERTAIN VARIABLES &. SENSITIVE UNCERTAINTIES
iUNCERTAINTIES CONSI:JEREI>
FOR DETERMINISTIC ANALYSIS
Forecast Load
Hydro Level
Western Energy Allocation In 2004
DSM Pefletration
Western Capacity Allocation in 2004
Natural Gas Pric"
DSM L"pact
Resource Availability & Cost
I'UNCERTAINTIES CONSIDEREI>
FOR PROBABILISTIC ANALYSIS
,Forecast Load
Hydro Level
Western Energy Allocation LD 2004
DSM Penetration
Having a manageable set of Ullceru,jr:.tits, staff men us.ed expen testimony and
analysis of historical data to assess the probahili ty dl stnbutions of each uncei'UUnfy.
Next. a detisioD tree was constructed with the dec:islOD Dode having five branches,
each representing & un~que alternative plan. Attached. to each decislon bnm<:"h were
the (-our LIllcerta.lnties, each having three states. Figure 1.4 shows the decision tree
l.l..S-ed (or the probabilistic p~J~e o( the analysis The probabilities of occurrence of
ea!:n uncertain outcome is labeled above that branch,
Qv.el"VltW
City ofPt:.w Afro 19911RP 10
..
'!I
"
•
,
A
!~-
DECISION TREE FOR PROBABILISTIC PHASE
DSM
Dei1etration
,4 C LJ-=---
\\ 0
1.3.2_6
(}y~rvje.
\~-
Figure 1.4
Results and rHommeadatioDs
The results of the probabilistic ana.:ysis can be pre!en~d as a curnuJative probability
rlIstribution curve. Figure 1..5 shows the cumulative-probability distributions ofPl~.';
A. B. C. D and E ;or SAR and Sc.
Figare 1.6 shows the expected val ue of tach of the pI ans and the extreme val ues for
both l .. eliZ«! SAlt and SC
From 'the above results, it w.!! wnduded that Plan B generated a high SAR and that
Plan E resulted in the highest SC. Therefore plans B and E were ehminated and the
impact of Plans A, C and 0 were further exa:m~ned to determine &hort-term and long.
term recommendsrions. These reco[11mendations are presented in Table I 4
City ofPak; Alto 199] lRP II
OVf.rvi,"W
CUMULA nVE PROBABILITY DISTRIBUTIO:\,
SOCIETAL COST
SYSTEM AVERAGE RATE
" r---__ ~----
-I
I
,~ t-------
t" t2~~~~~:·-~~~~~~-<,~:y-r:· 1==:: t 73 I • -----~p)r,C
~ C Q 0 Q cop c ~ Q 0 0 Q Q C C C ~
Cumulative prob.<lbility
----'--------
FigureJ.5
City (If Pa&J Alto 1992 18.P 12
,
•
•
•
O"el"llie,;<
~"~~~~~~:~:2: "
;;
EXPECfED AND EXTREME VALUES
I SOCIETAL COST I
I
I "" I~ , --------
51S .,
'8.'50
§ Itzj e -
!>:lO Ii'" ""-
"-""
",Z5 ""
400
A B
'l o.l,
C
FJ",
D E
_______ -.1
r SYSTEM AVERAGE RATE l 76 --
'J,Q
7S
10( "
r--~
"M
~ '-
-~ -E±
B
Va -"It
~--fC" -r····· "GH 1
.. tOW
tl~mf=t'
-oc
C
Pbn
-~ J
__ D ___ E _____ _
figure: 1.6
00' of Pal" A.lla 19921RP
-': .~:, ...... -,
<
13
Tabl. 1.4 RECOMMENDATIONS
SHORT-TERl\f ACTION PLAN
• Participate in the Seattle City Light Energy Exchange
-Do nOI panicipate in the NCPA STIa projeC't I
· Impl.men, pilo' DSM prog""'" with DSM lev.1 2 (30 MW) as the long-'.rm
Urge;
LONG-TERM FLEXIBLE PLAN
· Participate Ln sttJdy of Cdavera.s (French &: Ramsey) hydro enhancements
· Participate in st'..ldy of annual baseload gas proje.::t and swnmer-oni)" b~load
gas project
· Pumle 75% of DSM achievable potential (30 MW)
· Investigate local generation and co-generation options
].4 Olleniew at the Doeu menS
Chapter 2 discusses the load forecasts which fo['Ttl the basis for the lRP High. nomina.!
and low forecasts are discussed. along wirh the m<!thodology and assumptions used in
arriving at these forecas.s.
Owl"Vf".w
City "fPaloAlIo 1992lRP 14
•
/
Ch£pter 3 discusses Palo Alto's e-x..istir,g resourCe3, both dell1a...,d &rid supply, t."'tar were used
and con:.idered in this lRP, Descriptjuns of the resources are provide-d and 10ad/resource
balance chuts for existing resources and future resources are prOVIded.
Chapt~r 4 provides a description ~f the funlr, supply and OSM alremativ~, The screening
process, usea. in reducing the number of future supply and DSM alternati .. 'e5 to a -:cst
effective and mar.ag:eable set, is also described
Chapter 5 outlines the methodology used irl valuing envi.onmental externalities In this
IRP we considered only air pollution e~ernaJities.
Chapter 6. includes the d-esc;ription of me determLnistic analysis phase wherein the
appropriate leYel of eub of the resources descn bed in Chapter 4 was selected. This
description is followed by an explanation of !he probabilistic phase c;f the analysis VIIhereio
uncertainty was considered e~plicitly. Chapter 6 also includes further background 00 the
DA technique. Chapter 6 conclude'; this document with the results of the analysis and the
recommendatioos of the 1992lRP.
The appendices which follow the last chapter rootalns !Come of the detailed output of me
analysis and also • description of computer moJds. Appendjx. A provldes a thought
provoking dascussim] of the lessons learned in developing the 19921R.P and also includes
suggestions to improve the i994 IRP. Appendices B and C provi<k descriptions of two of
the .£PRl computer model used in zhis analysis. r.amely DSManager and MIDAS.
Appendix D oontair.s tables showing the detailed rerults of parts of the analysis.
Owrt1I~w
CiJ)! ofPQwA1UJ 19921RP IS
2. LOAD FORECAST
The load rorecast is a prediction oCme City's demand and energy requireme .. ts and is tl,e t-asis
for lR.P development The 20-year load forecast is 'Updated annuJl.!ly 8l'1d submitted to NCPA
(or transaction sclt~u1ing a."d t"Pense allocation and ttl the California Energy Commission
(CECJ for the development of the CEC bi"!:1I1Lal Electricity Report. b. addition, a short-term
for~a.c;t is developed ar.nuall)' for budgeting purposes
The 1992 fcrec5St projects a continuation of the decl~ning growth !'ates experienced in recent
years. Monthly peak: dem.u:d is ~pected to increase at an a .... erage annual rate of O,g percent
through the year 201 I, with allnlial energy ,sTO'Wth of an average of 0.81 percent per year. The
lower growth rates are targely due to mandatory conser.:atl<>n and diminishing buildotJt
opportunities.
The forecast is directly dependent upon the projections of several variables, 5uo:::h as weather,
price, energy usage intensity and total square footage being served. These variables are uncertaln
and subject 10 variation over time High and low forecasts were developed in an attempt to
identify the t-ounds of these uncenainties This section summarizes the methodology utilized in
developing th~ lead forecasts.1
2.1 F.reastinZ Methodolociet
The three most common met.~ods used in load foreca...qjng are trend. econometric, and
end·use. The trend melhod &ssumes that "things will continl!e on the path mat they have
reeentIy been on" and does. not explore the relationship of changing factors 10 electriclty
use. Th..!s method therefore tends to be aCoCurate only over the short term .and if nil
structural changes are anticipated.
The economelric method employs statistical analysis to dete-nnine the relarionMip ber-Neen
peak demand or energy consumption and explanatory variables such as price, weather and
economic activity, It requires !\ltw"e projections of e~:plU\atory variables, and a large dau.
base to handle historical data on demand., energy and explanatory variabJes UnlJke
trending. the econometric fOTe(:asting technique Cii11. respond to past or projected Lhange-s
in the influential factors and account for DSM impact.
The third method. end-use ana{vsis. examines the energy use p.ettems of each customer's
individual appliance or "end-use." End-u..c;e analysis can be apphed to any sector for
which the saturation and onergy use of each appliance is either known or assumed. End
use analysis enables utilities to m ore accurately esti mate dl e impact of changes in
A detlllled refereoce of the uswnptions" metlJodo[C'sies and data u!led to deye!op the foreca:rt .s. the
City of Falo ... Jto 1992 Electricity Report CFM'·9 prepared for the C.l,fornia Energy Commission
Lo4d FON(;'rut
City of Palo .4110 199) lRP
16
~
•
"
"
."
•
•
•
•
2.2
.. ,-"
-'-.-.--
ownership patterns.snd individual appliance efficiency. Thi~ met.~od requires extensive
data 01' buLltfing and appliance stocles.
HUCaritaJ EIec1ric:it)' Ule Bad FOre(utinl Method, URd iD PaJo Alto
Customer use patterns result in S1JbstantiaJ variatiOI\S in hourly electric: ~ad Generally
peak demand occurs during the day a."1d ~l1ly evening
Palo AHo's total electrical use, measwed by peak demand and energy consur.:lption, also
varies markedly by se&SOa and from weekday to weekend. The summer loMs are higher
than winter loads due to the impact of alr conditioning.
Electrical loads in Palo Alto.are categorized by end-uses such as air conditioning, fighting
and heating. and by custome,' sectors such!S residential, commercial and industrial. From
20 YEAR ENERGY FORECAST
BYSB:::TOR : F -,~~--:----~ ~-----I~--~~ -~-~
'" ~-..:-------
t .. 1--.--.-----.
~ 500 --------------. -------1: .. ----. .... . ..... _---
,.. .-.. c .................... .
r.::::= .. ...;;;,~I
I~ ....... I l_--:-··~I
1975 to 1983. total incrases in electricity consumption foHewed a relarivtly consistent
pattern. GroW!h rues for both demand and energy were approximately 3% per year, l'hi.s
pattern lent itself well ro !he forecasting method employ,"" by Palo Alto, that of trend
rorecasts "With judgement applied to adjust for expected future variations in trends In the
early ] 9S0's" Palo Alto began experiencing increasingly weather-sensitive load gr.owth in
the commercial and industrial sectors. Since the trend methoJ cannot oKCOWlt for weather
variations OJ' the impact of DSM programs, the City had beljurl to plan and implement,
• new forecasting appnl8.(:b, In 1983. staff began using an econometric. modeling
Lootl Fonc#IJI
CIty of palo Alto 199] IRP
17
-.
technique to forecast electricity consumption for all Pa.lo _"-ho customers
SLIt, aggregate econometric forecasting technIque could not distinguish among the various
custome~ sectors nor could it recognize the [u'u:e impact of mandatory C"on~r'\'ation
imposed by rtvl5ed building codes, To mitigate this deflciency, staff used a. cornbinatiC"n
of econometric and end .. use fQrecasting teehniques. The forecast of industrial customer
ioad was developed using econometric techniques. while forecasts of residential and
commercial loads w::re developed using end-use forecasting models,
Exami."\anon aflow by Cij..qomer sector in Figwe 2.1 reveals that the greatest contributor
10 Palo Alto's 1992 energy consumption is the comrnen;ial sector {62%) followed by the
industrial se1:tor (25%). Increases in total utility energy consumption ha.ve been driven
largely by growth in the commerdal sectors. Residf,ntial consumption, representing only
13% eof me low:!. has experien~d minimal growth.
2.3 1"% Laad Fore<am
The potentia: for future load growth in Palo Aho is somewhat limited. Over the next 20
years, the compound annual growth rate (CAGR) is 081% for energy and 0.8% for
system pe&k. Figures 2.2 and 2.3 show the 2D-year demand and energy foreG3StS. The
CAGR o{energy i. higher in the wly years oflbe fOfeeasted period (U7% from 1992-
96). The CAGR dectines to an average of 0.31% from 2006~2011. The decline in the
CAGR is attributed to reduced oppommlries to add floor space, strir.ge..nt land u.ote
planning criteria and the impact of mandatory applian-c:e and biJilding efficiency sta.,dards,
The en.;::(gy forecasts for the Residential, Commercial and Industrial seclOrs are shown in
Figure 2.], Currently tht commercial sector ~counts for 62% of the totai load and is
expected to grow at a compounded a.""l.nual rate of 0.91%. The growth is primarily nused
by an increase in energy usage ~ntensity (EUI) and a small increase in the commercial
floor space The industrial sector represents 25% of total consumptioD in 1992 and its
CAGP. is 0.-5&°/. over the 2.0·year fcr~asttd period Currently, the assembly industty
dominates collsumption in the industrial sector. In th~ future, it is likely that most
assembly industry will relocate to other cities in the Bay Area and w1l1 b.: replaced by
reseacch and development and offi,e space uses
The residential sector consumes 13% of the utility's total purchases. The residential
C'lns.umptioo CAGR is lowest {O.62%) over the forecasted period. This ~r is
categorizec! by a general trend to smal.ler fmnhes and a slow growth of housing units,
Frojected at 0.2% per year
Load FartcfUl
Clly of Palo .-tEte; 199] lRP
18
"
./'. ','
20 YEAR DEMAND FORECAST
Figure 2.2
20 YEAR ENERGY FORECAST
s ::f~= ---
~ IICOO') I -~
.~ loorot~ 7
! !0Cl0I.'" --- ---.
"' "'"
Forecast Year
Figur" 2.3
I.ooJ Fon~ast
City of PtJ0 Aito J 5192 IRP
,. ---Nominal i
--High I
: ---lAw i
.. '
19
2,4 Future Forecastinc Methods and EfforU
The Uriliti~ Department 'Will continu:aIJy expand its oata base and strive to impro .... e
forecasting rr.ethods. Emphasis wiJl be pla.:ed on the development of an end-use model
to forecast industrial sector etectri:::ity consumption, This is the fir.;t time Palo Alia has
forecasted energy use with a m!x. orelld-use a...,d econometric model!', Overall, the results
are satisfactory. There are, however, I few areas that need improvements One of these
areas is the sales data.. The square footage data should be updat~d and confirmed by
building type.
A more accura!e residential foreeast should be developed. Thi~ cl)uld be done by aitering
the CEC mooel or by usir.g a mooel better suited e Paio AltC"s usage patterns. EPRl's
Commend 3.2 should be in use by the next CEC fi:ing to improve the commercial
forecast and sho~ld include the mandatory conservation effect Forecast a.cc:.lra.cy should
also increase when more :nflJrmation regarding EUT's and fuelshares is available. End-use
models for the industrial sector ace not yet reliable due to the sector's small size and wlde
ranging div~rsi'!;}·. End-use models do ~ive reliabfe results for the re£idential and
commercial sectors tn Palo Alto, and as the inputs and mode!s improve, the confidence
Ie-vel will improve as well. Currently, load profile recorders ue being installed on various
building types in PaJ.;, Alto. Within two 10 three yean, there ",,;:11 be adequate data to
forecast demand for the sub·sectors of the commercial and industrial sectors
Load Forecast
Cily of Palo Allo /P9}lRP
20
J. EXISTING DEMAND-AND SUPPLY·SIDE RESOURCES
This chap!u offers I description of Palo Alto's existing energy resourcas, beginning on the
demand-side and ending up on the stipply-side rt also inc:ludes graphs of sea.<;()nal load and
resource balances.
3.1 Demand-Side Resourte!
Demand-Side Management (DSM) refers to ictentiona.! actions by the utility to influence
customer behavior and usage to achie've specific load shape objectives. The five basic
load shape objectives typically addressed through DSM are strategic conservation, peak
chppjng.loa.:! shifring. vaHey filling and strazegjc load growth. The eil)' of Palo Alto has
been an early propor.ent of DSM and has implemected. a number of programs directed at
strategic conse.~ .. ation. pe.al:: clipping and load shifting T'NO recen! electric DSM
programs are:
3.1.1 PARTNERS
The PARTNERS program was run from 1985 through 1990 to encourage
commercial and industri al cUS!omers to install energy effi ci ent devices that
reduced peak demand. The PARTNERS pro[Zt1lm was targeted II peak clipping.
strategic conservation and lo!1d shifting opportunities, Therefore. it has a relatively
small energy reduction compared to its demand reduction. ThLs program reduced
our peak demand by 6 MW and lowered our energy need by II ~ear.
The Load Managemea1 Program was initiated in 19&5. Participating customers
signed • Load Management Contract with the City. agreeing to reduce their
demand upon I telephoned request from the City The program's goal was to
reduce demand during pe.U. hours, thereby Ivoi ding the pur~hase of relatively
expensive PG&::E partial req1urements pcwer.
The customer benefitro from lower bills· the more power voluntarily saved, the
greater the cWotomer incenti.,.-e The City benefited by balancing supply and
demand,. thus avoiding expenslve supplemental power pl:1chases and lowering the
cost of power pro-duction.
biJrm, l>e...-md. attd '~lIppfy-Sid. Rrs.cnucu
CII)' of PIJI" ,f.ito 1092 lRP
21
,
3,2 Supply~Side RtsGurces
3,2.1 Wutern Are.i Powu Administnooill (W~.stern.) Contract
The majority of Pa!o Alto's electricny r.eeds ate provided by t.lte Western Area
Power Administration (Western). Western is the: federal agency that markets
power from the Central Valley Project (CVP) as wen 4.!: various Northwest
purcha!es and exchanges with PG&E. The present contract aHocates 175 MW to
Palo Alto from !964 fO 2004, Bl!\:ause Westem is expecred to continue to be a
r~lat:iveIy low cost resolJl'ce, maintaming the City's allocation is consider~d a top
priority. It is presu.rn.ed that Palo Alto will retain the majority of Its aIloctUian (95
tv-100 percent} follo ..... i.ng the year 2004. Any indications to the contrary will
become evident during the n".newaJ process of the Western contract,. 'Nhlch has
already begun. If Palo Aho 15 unable to retain more than 900/0 of its allocation,
resource planning activities and decisions wili be signific:andy impacted.
3.2,2 CalaveraJ If,..drotJe~tric: Project
In February 1982, the Federal Energy Regulatory Commission (PERC) granted the
Calaveras Projec:lli<:ense to the CalaverAS County Water District (CCWD). NCPA
financed il\d constructed the project under that license. Palo Alto has a 22.92%
participation level in the 230 MW Calaveras project At the 230 MW rating, the
proje...'"1 provides Palo Aho with 52,; MW and., for average watt! conditions, 110
GWb IOIlDUlilly (44 GWh ror dry year condition). Constructioo began in the spring
of 1985 and commercial o~ration began in February 1990.
Butd on an evaluation of the usability of the Calaveras Project. Staff ba!.
suc.;eeded iD laying off some of the project outp'Jt (6.52'%) to the City of
Rowvllle through the year 2004, leaving Palo Alto with a balanc. or n7 MW.
The portion laid oft' to Roseville will return to the City In 2005~ at which time tt
may either be ecotlomicaily LiSable tn Palo Alto or proposed to be laid off again.
This hydroelectric prcje;t will continue to supply some oi Pato Aho's peaking
requirements io the future.
3.2.3 Pacifi( Gu ud Electric orr·peak EDerv
In 1990. NCPA 'Uld PG&E settled a major dispute centering on a complaint
regarding transmission access which NCPA had filed with the-Federal Energy
Regulatory Cl)mmission. The resulting Settlement Agreement provided two
:services to NCPA. The first of these services provides NCPA with SO MW of
transmission capacity from the Department of Water Re.sources. NCPA may elect
to take up to 50 MW on a two-year commitment basis. This service may be
UiJR"6 D.-1If4Nl. and SlJpply-Sidt R.tJOt/rct;1
City of Polo Allo J 992 IRP
22
•
•
•
•
• ,
J.Z,4
withdrawn by either of the parties upon giving 24 months notice after the end of
199',. Therefore iile ~arliest withdr"w&i of this service is 1997. PaIo Alto is:
entitJed to .approximately 16.5 MW of litis service, The sec.or.d service provides
for 50 MVl cfoff-pea.!: finn energy rnml PG&E In January of 1991, the City of
Palo Alto laid off' 2,S!tw{W of its entitlement in the PG&E off-peak energy to the
Cil'j of RoseviJle. This leaves Palo A['o with 14.0 MW of PG&E off-peal
me!:'&)"
Wasbin&ton Water Power CDntract
NCPA has contracted with Washington Waler Power a.-::d the Bonneville Power
Administration to purchase' S 0 ~(W of Ii rm resource. Reso urc:e delivery depends
on me 'su,ccess!-ul completion and operation of the C&iifomia Oregon Transmission
Project (COTP). The contract i. ",heduJed to be«>me effective ill 1993 0' when
the COTP is in oper1ltJon, "itichever occurs first Palo Alto has an 1],3 MW
et'Ititl~ea.t in this rescur~e, In early 1991, Palo Alto agreed 10 assign. at cost..s
MW of this resource 10 the City of Roseville, Tbts portion 'Will b.e returned Ie
Palo AI", in 200S.
3.2.5 PG&.E Portia! Reqolreme.1
If th~ City is l!1Iable to meet its load requirements through its exlsting resowces
or economical purchases from third ~arties, then PlI1!iaJ Requirements power may
be purchased from PG&E through provisions cootaine<l ill the NCPAlPG&E
Interconnection Agreement Palo Alto relied on a very smaJl amount of Partfal
Requir~ent power in several swnmer months of 1983. Betw~ 1984 md 1990,
when the Calaveras Project C&mC on lin~. Palo Alto made several cost effecti\'e
sho:t·term pW'chases from suppliers other than PG&:E to supplement its Western
resource, -:bus avoiding the use of this more costly resource: .
.3 .2. 6 Calirom ia-O recoa. TransmissioD Pmjeet
The Cahfornia.-Oregon rJarumi~oD Project (COTP) is & transmission line that
~x.tends from Q1mda subs-tatica in Southern Oregon 10 Tesla substa.:ion in Central
California. The COTP inc, ...... the transfer capacity from the Pacific N,nhwest
by ) 600 MW. Construction ~f COIP is well \mderway with the Tr.ensmission
Agency of Northern Califc:nia (TANC). of which Palo Alto is • member, .cting
as the Project Manager. Other members of T ANC include West-=m. SMtrD. the
Cities o(San .. CI ... o, Re<lding and othe, NCPA members plus the City o( Vernoo
and tho Modes!<> and Turlock lmgation Districts Th. COTP Will provide
transmissio-n ~«Ss to the Northwest 'Nhich would ottterwise not be available to
members of th~ proj ect. making possible the purchase of reI atively inexpensive
Ui1fl'fg DtmanJ· 41Id S,.pply-Si.h RtJ01lrtU
City 0/ Palo AltCl 1991 lRP
23
Northwest power and allowing members to
exchanges and 0 ther economical transactions
project is approximate!y 50 MW.
er.ter agl·eements fc,r seasonal
Pale Alto's present sha~e of the
In early 1991, Palo Alto agre.ed to assign 7,68 MW of lts entitlement in the project
to the City of Roseville in conjunction 'With a concurrent layoff of ~ M\\-' of Palo
ALto'!! share of the WWP contract Tnis portion will be returned to Palo AJto In
2005"
Figures 3.1and 3.2 show the physical baJance between forecasted demand and energy and
tht capacity and energy avaifabte from existing resources arod the re50u,n:,es recommended
in P!an C (see Tw.Ie l.2) for summer and Mnter. DSM implemented prier to 1992 has
been included in the forec8bt of demand and energy. Figwe 3.1 shows that Palo Alto is
capacity ricb in winter in comparison to summer. A similar situation is exhibited 00 the
energy front. Figure 3.2 shows the load and re~urce balance if all the resources
enumerated in Plan C are implemented.. The additional resources compensate for the
shortfall of upacity IUld energy whicb would occur If we did nothing and the fuiUre
unfo!ded 8CU)riling to our fore<:a~t
EXi,ti,.. D,,,,(iN/-and Supply-Sid~ XtJavrcrr
CifY "fPakJ Alto 19921RP
24
•
•
,
,
•
]
]
• I
i r
~
r----l
I J i
I ~ J < !' lil~rl ______ J
•
E.l:ijti,,~ DrmrznJ· and Sllpply-Sid, RUQUN:t!J
City of Paw Alto 1992 fRP
no<
'''' """ "'" "'" ...
"'" ...
"'" too< I ,00<>
"'" .,"
, 166!
•
•
"" "'" ... '
""' t ... ,.;"
"'" "'" """ "'" "'" -"'" "'"' "'" "" I ""t> """ ... , ....
"" "" ..-, .,"
"'" ""
"''''
'--rl
I ~ d ~ J
LUl~L
25
1
1 •
. .
1---)'
I !tJ'& "II I ~ I! a 1 ~ [ ! ! I
ltUlBJl1 U 1 i
ExulinK DtmaNi-a"a Supply-Sidt Ruc/.orcrr
Cit}' of Palo Alto 1992 /RP
"'" "" <ox
""" "." '"" "'" "'" "." =, [00.:: ,..
""' 066' ... ,
"" "'" "" ..."
"" ""'
I -.----• I
I f],
\ ' 'l" i '. Ili1JJJ[.h
II i ~ ~ ; I I ~ I
~---------~ -------------------
•
""
"" ".,
26
, .. r
.~ ,
'!' ..
•
•
•
•
•
•
•
4. SUPPLY AND DSM RESOURCE SCREENING
This secCou describes the supply :md demand 1esour~es that were ronsidered as future
alternatives and the process we used to screen these resources Into I manageable shortlist
Information about the supply resources were gathered mainly from rhe responSd to NCPA's
Request Cor Proposals (RFP) received in N ovem~r 199], A description of the types of supply
resources is presented and then followed by the analysis used to screen the resources
Information cn the demand resources was gathered mainly from the 1992 DSM report, which
contains a detailed analysis of' the ter.-hnical and achievable potentials of D; varlet)' of eneig)'
efficiency measures in Pale Alto, This chapter describes the methods used to 3Creen these
measures and ihe procl!'SS of oonsoiidatin.e cost-effective mea'iwes into dlfferent programs_ rne
programs were then arrayed as thre.e o,,"erlapping "levels: each higher ievel being a sllp~rset of
tho lower one(s).
4.1
4.1.1
Sapply-lide resOUI'US
In the summer of 1991, NCPA issued a RFP {or power sl.lpply projects or contracts and
third party energy ~ci-=n~ programs. The responses Ii} this RFP served as the main
information database. %0 whidl we added • few mo~ resources prior 10 screening.
Supply-Side Resource Sc:reenin, for IRP
NCPA re<eived 84 supply-sid. proposals to .. "ng 7.321 MW. They. then,
included 4 NCPA projects totaling 207 MW for a total of 7.528 MW. All 88
options were IMltereci inrc a Proposll Eval uation Model (PEM) by He.rno\,'OOd
Energy Services Inc. (HESl). E!Ch proposal was modoled against • mark.t
alternative to determine its benefil cost ratio. The costs and benefits were then
adjosted to reflect the effects of front loading of 00"". timing of proposed start
date. technological or fuel diversity of propose! compared I<> NCPA', existing
resowce mix and the feasibility of the resource getting permitted, financed md.
built The resources wne also given sizing penalties if they were 100 small to
bother with or too large to easily tit into NCPA's !e!oCiW"ce mix. NCPA chose to
use the California Energy Commission'., air pollution n:temality values attrib,.ited
to each resource type. HESI escalated T~e va.iue of tho~ externalities at the 4.S%
inflation rate.
Based on the adjusted DIC ratios. NCPA screened the U proposals doW!> to 21
resources with I total of 1,605 MW. Thi. included 1 NCPA projects with I
combineJ capatity of 159 MW and 18 third party resoW'(:~ with !. combined
capacity of 1.445 MW.
Palo Alto staff also ""ded two 2 MW fuel cell projects to the list They are the
early producriC'D unit and the commercial unit of the molten CI.(bonate {uel tells
Sllppiy DNI DSM /UWlUCII &n."i:n¥
City of Pedo AlkJ 1992 1RP 27
slated for delivery in the 1999 to 2002 time fr&me.
Tabl. 4 I NCPA SHORTLIST
I NUMBER Of PROPOSAlS t CAPACITY
TYPE (MW)
Coal C-as Hydro System
In •• rmodi II. ruspllchab 1. 0 8 0 0 493
Baseloa.l -Ann!J.II I 3 0 I 521
Baseload • Summer only I 2 I 0 0 386
Power e~t:hange 0 0 0 I SO
Peaking -Annual 0 0 2 1 151
Peaking· Summer only 0 1 0 0 4
Total 3 JJ 2 J ]605
Palo Alto staff received a copy of the PEM and the: data base of proposals in
Marcb 1992. Staff customized severa! criteria in the adjustments ro better reflect
Falo Alto's unique perspective, The custorniulian included reducing Lite small
size project penalty threshold to reflect Pall) Alto's wilhngness to ac;quire small
resources, altering the resource diversity penalties to reflect Palo Alto's current
resource mix and chMging the escalation rate applied to externalities from 'the
general inflation rate to the discount rate {so the real cost of e)(temf\]ltjes would
he constant over rime).
All 23 proposal. were then run through liES!'. modlfiod PEM to calculate BIC
ratios specific t·..) Palo Alto, The resources were grouped by t':!e following types ~
seasonal exchanges. baseload. summer baseload. intermediate. peaking and
summer peaking.
In tl10se groups., the proposals \\lith the highest adjusted B/C ratios were selected.
yielding the Seattle Clty Light seasonal exch3C!ge. the Ke:n River baseload gas
fired repowering, and the AES Gray's Harbor baselaad gas-fired 5urnmer-only
resource
S"pply and DSM R,,(}lI1'Ct SUttl1i",
City 0/ PaJo Alto 199] Irq. 28
!
,
I
• The french Meadow and Ramsey Calaveras hydra enhancements were ,hosen
becam:e they wert the high~1 rank:ing pea!c:ing units:, they ~ave a high degree of
controllabi!ity, Palo Alto Wll1likely face .. ~nd phase part:i<:ipaboc decision in
the next year 00: Lhese units. The NCPA SnG project was also Included since it
wac; of particular interest in the plan. The ;;ot:1mercial molten carbonMe fuel cell
was chosen because it was the hiehest ranking locally sitable !liupply resource
The table below sho'IWS several key a.ttnhlJtes of Palo Alto's shonlisted sup?1)'
l'esourc:es.
Tab!. 4.2 PALO ALTO SHORTLIST
Percent Percent Real
Fixed Variable Melded EXler. Total
MW y .... CF Cost Cost Mills Mi!ls MIlls
(I) (2) Reso= (3) (4) (5) (6) (7) (8)
0,5 1995 STIG 15% 24% 76~. 46.3 6.6 52.9
0-15 1993 Exchange 5W. 10.9
0-10 1998 Baseload 8'/"1o 100010 0% 29.6 6.7 363
0-10 1998 Summer 95% 96·1'. 0% 40.9 10.9 5U
Base
O-H 2000 Freoc:b 4~;' 100% 0% 56.2 0 56.2
0-12 2002 Ramsey 31% 100% 0% 67.4 {) 67.4
0,2 2000 Fuel Cell 95% 34% 66% 40.8 5.1 459
(t) MVr" eumined
(2) Start dau \)( re~'..lroe
(J) Annual of seasonal capacity ractor of rnource
(') F,"tion of cnm thai do !lot "'41)' v-ith usage
(S) Fraction. or costs thai "'&1)' with usage
(6) Ddlated am:w&l ~ divided by IlDDUliI output of 4st~d CF
(1) ExtmIIlity CO'SU ~ in mills/K.wh
(8) ReaI melded mills pillS ex:temality mills
Th~ DSM resources fall into two major catc-gones, those proposed by Energy Senice
Companies (ESCO) vi. the NCPA RFP, and rhose thai would be developed i. house by
~pply dItd DSM liuovrc~ ~r~,"j"6
City ajPawAlt<; J992 lRP 29
Palo Alto staff. Twenty·four demar.d-side propc.saJi totalling 139 MVl were reaLved.
in respocs.e to the RFP, Unlike the supply·side resowce s<:reening. the majority of
in.forrnation about DSM resources was. developed In-house: .
.... 2, I Pnlimill.r, Stl"Hninc of NCPA RFP 9ropo!l.b
The twenty-four dem81id-side pro~s leceived were first evaluated ba.se<! on
pricing USirlg SePA', Proposal Evaluatioll Model (pEM), The PEM compares the
cost of a proposal (Including customer costs) 'With the avoided market purchase
and calculates a benefit-<ost ratio (SIC) based on the total resource cos;: ("IRe)
perspec:ti ve.
In ldditiOil to calculating th, TRC. the other attributes NCP\. considered in the
screenlng process were the marketing plan, savings venfication p!atJ.. persistence
of'savings. technology, financing and experience, These factors were included as
weights in the calculation of an adjusted B/C. A DSM proposal with weakness
in plan.'1.ing. verification or marketing would have a lower a.djusted B/C compared
to a well-planned DSM proposal. NCPA then evaluated the proposal:ii based 00
the above data and decided to: eliminate proposals with an adj'.lSted SiC below
1.20.
This prelimiOluy screening decreased the Dumber of DSM proposals from the
twenty~fOUl proposed initiaHy to eleven proposals totaJiing 53 MW. A few
pr.;)posaIs overlap since some Eaergy Service Companies (ESCO's) proposed
several pricing 51nictures. The competitive DSM proposals included J lighting,
2 HV AC, 1 motor -and 5 general DSM programs .
•. 2.~ Second S.....,.iD' of NCPA RFP PropMalt
Two of the remaining eleven proposals could not be adapted to Palo Alto's needs.
The other nine ~'ere then screened 'W'ith a detailed DSM evaluation model
(DSManager ~ description provided in Appendix BJ, set up to mimic Palo Alto's
system and the syecifLc. costs and benefLts to Palo Alto for each lnd:i~dual
proposal.
This S4;r3ening resulted in benefit~st ratios for each proposal from the TRC and
.impact on rates (RIM) perspective. The five proposals, with e TRC B/C above
1 (S .. Table 4.3) were then ,valuated based 00 TRC. RIM. and the end-uses
offected.
SIIPply aNi DSM RUDW'C_ Seraninz
Ciry of Pede A.1.o J 99 ~ IRP 30
•
iI
•
• •
• ,
,
•
Tabl.4.3 NCPA DSM PROPOSAL SHO;tTLlST
PROPOSAL I RIM ITRC I Annual GWh 1~:~alMj .... "
New Construction 1.42 2.68 23 1.0 I
Lighting I 0.69 1.07 9.3 1.7
Lig.\ting 2 0.90 1.84 40 0.7
Lighting 1 0.18 138 4.0 0.7
I HVAC 1.32 2.20 1.6 O.S ---.J
The recommendation (rom the second screening was 10 punue the best lighting
(Lighting 2) program .... d the new construction program .
Since the completion of the second screening. several ESCO's have-submitted re.
bids. These bav. been modeled in DSManager as "NOli
4.2.1 Developmeat or io·bouJe proaranu:
The 1992 Ele<:tric DSM Report bwlt the foundation for the in-bouse DSM
?rograms. This report ev&!uated ov~r 140 different DSM measures and estimated
the technicaI and achievable efficiency potential of tho~ measures in PaIo Alto
The measures were 81'ouped into eight commerci alfmdustri a1 programs and four
residential programs. Table 4,4 lists the twelve programs Uld die end-uses
affected. The program description included panicipatiOD TOiles. fr~ ri ders, and
rebate levels Cor eadl measu...'"e in the program. These numbers were based Oil staff
opiDion md outside research put into each measure in ttl e DSM report. In
&ddition, tb.e program descriptio,. induded OutlllYs for capita! and administrati .... e
costs needed to run each program.
SIJIIJlIy DIt4 DSM &JOtUt:. Serum"l
City of Palo Alt.o 1992 IRP 31
Table 4.4 . IN-HOUSE DSM PROGRAMS
COMM£RCIAUINDUSTRL"L
Indoor Tube Li shting ROB
Indoor Bulb Lighting ROB
Indoor Li ghti og RETRO
Outdoor Lighting ROB
HVAC ROB
INAC RETRO
Refrigeration ROB
Refrigtration RETRO
RES ID ENTIAL
Lighting ROB
Heating RETRO
Appliances ROB
Water Heatinll REnO
"ROB
" RETKO
Replacement on burnout
Retrofit
I
The DSM report was also used to obtain data on ann uaI KWb saved per
parti cipant, incremental cost of a measure~ and the life of a measure. Th~ lives
of the measures were decreased to account for degradation due to '!arty retirement
whi cb occurs if & ;;usto mer is not satisfi ed with a measure, if the customer moves,
or ;f the building is renovated or tom down.
After the programs were defined,. load shapes (or each program wert developed.
Since Palo Alto performs only very hmited load monitoring of individual buildings
and does not bave CU\y appliance load shspe metering programs, most of the load
shapes were ob tained from outside sources. The largest share of the load shapes
came from the California EneriJ C.:.mmissio'1 (CEC) ""ito actuaHy obtained the
data from PG&E service are:as surrounding.. and therefore judged similar to, Palo
Alw. Residentia.:l lighting and heating load shape e:;tlillates came from Electric
Power Software (EPS), the consulting firm who designed the aforementioned
SNPply and DSM &.JC1IU"cr ScrunJ"g
Cit)' o{PlJioAlro 1992 IRP 32
l
)
,
42,4
nSManager evaluation model The only load snape based on acruaJ Palo Atlo
data was thai for the ccmmerciallindustrial HY AC programs
Modelinr o( in-houH proJrams
The data for each individuaJ program was then entered into the DSManaa;er modeL
To predict the impact teo Palo ~ .. lto. the model was set up with PaJo Alto's hourly
system load. rate and rate c:!ass desc~pticns. hourly ma..-gi."lal ~mergy cost" .Il1d
annual avoi.ded capacity costs. In additior.. data such as transmiss[on and
distribution losses Illd costs, tax rates., discount rates end deprcci&tion schedules
w(:re included
AaaJyria: or ja-bou. prolramJ
Each DSM program was run through the model to calculate B/e ratios based on
1h. Total Res<mce Cost (TItC) and th.lmpact on Rates (RIM) perspective. Table
4.S lists Ihe twelve programs and their lISS(X;iated TRC BIC and RIM SIC All
1he prognuns passed lbe TRC SIC tosI (TRC BIC above I " whil. the resic!ential
programs and lbe col!lmercialr.ndustrial outdoor lighting program bed • RIM SIC
below 1. A RIM BIC below ] could be improved by altering rebate !~eI5 or by
offering other financtns opti.-:>ns, though these actions might affect participation
rates as well.
Supply oM DSM JU~. Screertl"6
City of Pi210 Alto 1992 lRP 33
Tobie 4.5 COST-BENEFIT OF INDIVIDUAL PAl.O ALTO PROGRAMS
Pal. All. DSM Prolram. RL\I TRC Annuar Annual
GWb MW
.saved !leVIed
Residential Lighting, ROB 0.80 3.18 8.2 0.9
Residential Applian< .. , ROB 067 1.69 47 0.7
Commercial Lighting, Tube. ROB 1.39 423 35.9 6.7
Commercial HVAC. ROB 2.00 3.78 371 1\9
Commercial L,ghtins, BulblSocke~ ROB 1.57 3.28 493 9.2
Commercial R~frigeration. ROB 1.29 2.33 5.7 0.7
Commercial Refrigeration, RETRO 1.05 223 2.4 0.3
Commert:ia! HVAC. RETRO )40 L16 OS O.l
Resi dential Hesring, RETRO 0.53 2.32 2.0 0.03
Kosi dential W.",heat, RETRO 0.65 203 03 o OJ
Commercial Lighting, Outdoor, ROB 0.79 1.30 IS.7 007
! Commercial Lighting, Tube, REnO 1.20 203 14.9 2.80
The next step was to deterrillne bow to analyze DSM programs using the MIDAS model.
The pmgraIDS were ranked based on the IRe SIC and then combined into three !evels
of DSM participation, taking into consideration not only TRe BiC, but also customer
class, "'...ase of implementation and size of each level The resulting three levels are shoVwTI
in Tobie 46
SllPPIy aNI DSM RlICtU(, &runirlll
City ojPtJIa A 11.0 H192!RF 34
i
'-::." " •
Tab1.46 PALO AL'IO DSM LEYELS
rSM LEVEL RIM TRC Annual GWh Annual M"~,' l
saved ~ve~
LEVEL I 147 370 84 20
7" of AchieWlble Potentia!
Residential Lighting, ROe
Residential Appliances, ROB
CommerciRl Lighting, Tube, ROB
Commercial HVAC, ROB
LEVEL 2 L47 3.40 140 30
78% of AchiellClble. Potential
~; as Levell, plws:
f:ornmercial Lighting, BulblSocket. ROB
~mmerciaJ Refrigeration, ROB
Commercia! Refrigeration, REnO
Commer<ia.i HV AC. RETRO
~EVEL 3 1.36 2.9j 171 33
~5% of Acllievabl~ Potenrial
fsame as u Vt;/ 2 plus
"-esidential Heat 'og. RETRO
~esidential Wltemeat, RETRO
~ommercial Lighting, Outdoor, ROB
Commercial Lighting, Tube, RETRO
Level I includes both residentiai and commerciallindustriaJ progra.-ns and if implernent~d
will save up 10 4']9/. of the achievable efficiency potential esti mated in the DSM Repon.
It includes programs that can be offere"'l to most customers and is fairly easy to
implement Lev~1 2 adds four commerciallindustrial progn!.m5, with 78% of the
achievable potentiaL Level:3 includes all the proposed programs and can save up to 95%
of the DSM reporfs achievable potential
These three levels were model ed in D SManager and, as expected, all three passed both
the TRC BIC and RIM B/C test (see Table 4 6) The three level. were thee .. poned to
MIDAS to compete again5t the supply resoul'c!;S.
SJiPply and DSJ,{ Rumoret Scutl1mg
Cl"1y of Pa-to Alro J 99] lRP 3S
,
-f.
5. ENVIRONMENTAL EFFECTS
Environment&l effects: are incorporated in the seleclion of resources by including societal cost into
th~ decision making process.. Socie-taI cost, in its broad defjn~tion. includes the effect of energy
usage on society as a Vw-hole This indudes the effects on $O('ial. ~nomLC and environmenta.l
systems. The determinatton of socie.a1 costs of energy usage in this RFP [s restricted to the
qUDrltification of environmental impacts that are not accounted for in the price paId by the utility
Of its ciatOmers for eflergy produced or consumed These unaccounted for costs are called
envi.onmental er.ema!ities.
In general. environmental externalities may includ~ impact an [and and water systems Palo Alto
bas narrowed the 5(".Ope of environmental c:o<temaliries considen'd in the 1992 IRP to addt~s oniy
air pollutiM.
While several states have mandated the expHcil consideration of environmental externaltties in
planning for resource acquisition. neither the CaJifomia Energy Commission nor the California
Public Utilities Commission have forced the !ssue in California. Ther-efoo'e, Palo . .o\ho has the
opportunity to examine environmental externalities and to develop its subjective qUoUltifintion
of environmental impact
This S«tlOQ is an attempt to define difTerenr type;;: of eQvironmental externalities and the r~ns
to consider tbe~ roUowed by & brief section on some methods of valuing environmental
externalities, and a description of the values ~ by Palo Alto in assigning a cost to air pollution
enemalities.
S.I Reuonl to cOlllider envireoamentaJ e-rr~t.s iD planniDI
Usage of aU energy sources generate environmental impact mal may not be reflected in
the price paid for memo Table S.I provides examptes of environmental impacts associated
with etectr(c generation from different fuel types. It includes impacts that were nOI
considered in developing this IRP.
E7JYlI'omfuntai Fffuu
Cil) of Plllo .4.110 199]lJi.P J6
•
•
i
" ..
.. -
Tabl, 5 I IMPACT ASSOCIATED WITH ELECTRICITY
GENERATION FROIII DIFFERENT FUEL TYPES
FUEL TYPE UPSTREAM LI\{PACT DOWNSTREAM IMPACT
Coal Mmi ng & surface landflllsiash disposal
rectamatlon. Enhanc;ed eli mate change
Acid ram
Slung impacts
Oil Drilling E!lhanced climate change I Pipelines Acid ~ain
Offshore Channe[s I Slting impacts
Tanker spills
Nstural Gas Drilling Enhanced climate cbange
Pipelines Siting imp .... "tS
Nuclur Mining & swfa.c:e W ~e disposal
rectamation Accidents
Siting impacts
Hydro Erosion Siting irnpacts
Flooding
Siltation of stream b.d
Fish impacts i Recreation i mpaC1.s
TraditionalJy utilities made decisions about resource acquisitions based on t~"'t economic
cost. without regard to the impact on the global environment. The underlying assumption
was that by complying with alt federal and stste environmenw laws. all appropriate costs
were internalized_ By !'lot recognizing the impacts to the physi..-:.al and bumlill environment
this assumptior' effecti ... ·e!y placed I uro vatue on envlrcIlmenta.l extemahties.
Though it is cfifficwt to pl&.ce a pre<:ise value on environmental externalities. we operate
under the .assumption 1ha2 i1 is non-zero Palo Alto's strategic plan has I goal of
providing e1ectricity.in an environmentally sound manner. By including a non-zero value
for environmental externalities Palo Alto LS bener able to assess the environmental effects
of its options and factor those effects into its planning decisions
EIf1>;"i»'IJIJf,"lDl Effuu
City of fakJ Alto 19iJ2 lRP
52 VaJ'Uidl environmental ntem.litiu
Thert we two basic approaches If) ca!culating envu-onmenta1 externalities associated wit."
s:r poUution: one is ~mmonly called dinel damuge e!limalion, whi1e the other is ~alled
revealed preferences I.
Direct ciam.age tstimario~ atrempts [0 eva.!uate the damages that can be direcdy linked
to the emis-;ions tlf I particular pol1uW'lt This approacb. though intuitively appeolling
suffers the burden ofincreastd c~mplex.jTy Direct damage estimations involve the direct
estima...~on of the impact on the mviro!1m~lIt and human health. Such estimates involve
identifying and qUiUitifying the &mowt of each pollutant emi"l.1ec, de~ermir.ing how each
is transported or dispersed. into the environment, determining the populations that are
exposed and the ex.pot3.lJn: respon.se, and finally. determinirtg the cost of e)(J>Osure.
Revealed J,lreference approaches use potl'JtiOD -control COSlS 1n1posed by regut&tOry
decisions as a proxy for the true externality costs imposed by • polhJlant. This approa(.h
is also called "control COst" or "shadow price" technique. This method 15 used because
it!s easy to ulculate. though the results produced from this technique may vary widely
from that ohtained through direct damage esti mation techni ques. This apprClach 150: ,roes
that regulations actually embody society's value of pollution control. The problelJ with
this apyroach is that society's pr~fe.ences change and a previous will ingness to pay may
nol imply a willingness: to pay in the future. The abatement costs that ere used in this
approach are determined by using the best altai lahle control technology presently available
to reduc;:e marginal emission.
In Electricity Report 90 the California Energy Commission {CEC) Staff used a revealed
preference analysis to estimate the vaJ\J.e of &..ir emissions reductions of SO:. NO •• CO~.
!tOG, and porticuJale$ smaller than 10 microns (PMIO), taking the data from Southern
Cal ifomia Edison territory and the South Coast Basin in Los Angeles.
5.3 E1f<rnaJity nl.n uoed by Palo Alto
At an ea.rty stage of the IRP process we proposed using CEC values for eX1e"llaiities.
These values are presel'lte.d in Table S.2. This was a compromi5e so~utiO[l since ..... e have
fiO wa.y of knowin& at this stage. till: value that the dtizens of Palo A:to place on
environmental external iries. In our next lRF we wi II attempt to -code-the externality
values of the citizens of Pal~ Alto. This is I very complex: endeavor and we will
approach it in a prudent manner chat aims for an equitabte solution.
--------
IThere iJ. • van body of litctahae dealmg with appro.a.cbu \0 valuing n!erna.utie1. The di~us~ion bere LlI
centered on the material ill ·ClJIIfptJr-o,i.n: Anaiy.riJ of MontUJI'J! E.rh",ar.J of Edunal Co.rLf As.rocraltd .... ie.
Combu.ttlOOI of Fonif F_!.J~, • ptlpe1" alnhoreO. by k>"M.tban Kooroey of La~ Berk.eiey L&bora"loI). in
January' ]990,
EnviI'-anm~ntal Effects
01)1 !l{Pai.:J .141m }99] lRP 38
• • •
•
•
I
"
Table n EXTERNALITIES FROM NEW POWER I'LANTS (1989
CENTSIKWH) DEVELOPE!) BY CtC STAFF USING
REVEALED PREFERENCO:S
Combined Combustion Combined Com busticJJ
C,..dr Turbine-Cycle Turbioe
Gas Gas Oil Oil
Heat Rate !440 13900 &440 IJ900
(Bru/Kwh)
S02 0.00 0.00 103 1.69
NO. 104 1.11 104 l.71
CO2 0.35 0.57 o 5J 0.&4
Total 139 1.69 2.60 4.25
EPJ1Ii,.~r"UzI Efftc!J
City of Palo .4110 1992 lRP
..
Coal
9660
3.53
3.56 • I
0.76
7.85 !
39
6. RESOURCE INTEGRATION AND RESULTS
This section describes the process of integrating the supply and DSM resources ... '1at remained
under collsideration after the static screening processes described in Chapter 4 The shortlisted
supply and DSM resources were integrated. in two stages. The first stage was limited to a
detenr.inistic analysis and the second stage involved a probabilistic analy~is, The deterministic
analysis helped us develop a reasonable number of credible altemat;e ptans, each being a unique
combinatioD of supply and DSM resourCes, and also helped to shortlist the U1lc.ertain voriahtes
to those that were most sensitive. The prob2.bilisric analysis $.t2.ge assessed probahilir,es of the
sensitive uncertain variables !hrough a combtnation of structured interviews of experts and
examination of historical data:. A dec.:.isioll tree was constructed (using !t.11DAS) describmg the
different plans and the uncertainties eifC(:ting these pIMS. The resu[ts of the slmulation conclude
thlS section.
6.1 Dettnniaistic Anafysil
Entering :hi.!! ~e of the process, we had a short1ist of viable supply-side and DSM
al'tematives, desc:ri~ in Chapter 4 in Tables 4.2 cmd 4.6. We had fonnulated the
problem.. idenbfied the list of uncertainties mat could impact our decision ant:! established
the criteria used to evaluate analysis results, Further, we grouped DSM options intO three
J,-;ve1s: or programs, eacb representing I. percentage of DSM acbievable potential over the
nut 20 years. The supply¥side resources were not :as clearly defined at this time.
Though we bad identified potential supply rescurces, we had used.;. screening model that
could not indicate the appropriate size of these resources to meet Palo Alto's nEtds.
Therefore, the first task was to determine the appropriate size of each of the shortlisted
supply resources,
6_ U Ettabll.hiDl tho size of .~ppl,. resoUr'Cfl
The first step in the deterministic analysis phase was to set the upper and lower
limits of each shorthsted supply resource.
To detennine the upper lL.'td lower limit on the size of the supply-side resources,
each was evaluated in combination with the STIG resource and each of the three
DSM levels.. Though we decided to use so.:ietal ~1 as ... criterion., the lack of
data Olll emmalities at this stage precluded using it.. Instead, we used roW
resource oost (societal cost excluding ex.ternalities). The structure of the tree used
to perform the analysis i5: shown in Figure 6.1. This tree was used to dtst-ermine,
one supply resource at a time, the the appropriate siu boundary of each of the
supply resources.
Rf!UJ'/Uc. i1lUSf'01fOtl "" Rllwlt.J
City of Palo AlttJ JP91lRP
40
-.-•
•
I
DECISION TREE TO ESTABLISH SUPPLY RESOURCE SIZE
Go (
r-(----~lL_
~ No.go '-\ I" _.,
---\ [..eye! ~ ________ _
S bartl lsted
supply resQurce
State I (Upper limil)
j State 2 I,~
II State J
\ State 4
State 5 (l..<M.u l:imj!)
Figure 61
--<J
the de<;ision tree of Figure 6.1 sholN'S the various combinations of' STIG, DSM
and suppry resource!:. The right-most decision node of FIgure 6.1 has five
branches, each branch refers to a unique size of the supply resource For exampte,
when analyzing the base g&S resource, we set the upper limit to 10:MW and the
lower limit to 2 l\!W. There were" levels in between the upper and lower limit
of 4 MVl, 6.MW and 8 MW, respectively. On the lett-most node of this tree, we
modeled SnG as a two-branch tree. The branch labeTed "Go" refr:rred to the S
]."fW level and me branch labeled "No Go" ieferred to no panicipation in the STIG
proj eel
This tree was modeled in MIDAS and the analysis was performed for each of the
six supply resources Tabl= 6.1 ~ummarizes th: results of this analysis phase,
showing upper and lower limits of each of the six suppty .resources. One point
to note. this method does not .ensure optimal level selection. Never-..heless.
choosing bou.'ldaries for each supply resource, ensures thai the optimal level lies
between these limits.
RnQurr:t lrllJITQtirm &-&slllu
eil)' of P .. lo 11.110 1992 /RP
41
·-
Tabl.6.1 UPPER &; LOWER BOUNDARIES OF SUPPLY RESOURCES
RESOURCE STATE (MW)
Uprer I Lower
SCL Exchange 12 6
Baseloarl gas 6 0
SummerooOnly baseload gas 6 0
Frerl(:b 3 0
RBmsey 3 0
Fuel Cell 2 0
0.2 Determininc the five alura.rive plAllt
6.2.1 Aualyzinl an cambin..doDI .r rupply l'HOurca and DSM levers
The process described in Section 6.1, I resulted in size identification of e-.ach of the
si~ supply re50ilJ'ce5. The procedure described in Seerion 4.2.5 rudted in three
cumuilltive DSM levels. USing these results., we were in a position '0 evaluate
combinatJ~ cfsupply and DSM resources simultaneously and exhaustively. The
different tornbina!ions thm: were evaluated are represented graphically in !:he
decision tree shown in figure 6 2
this tree has. 384 unique combinations of the sevea supply resources (the six
sllortlisted plus STIG) L"1d three DSM levels. For example, by moving along the
top branches of aacb Ilode. we are e\'aJuari:ng a pian which considers every supPlY
resource set to its upper limit and the DSM resource at its Io~r limil
Conversely. by scannin8 the bottom brancb of each node. we owJine a plan which
seb each supply resource at its minimum and the DSM Tesource at its upper limit.
Between these twn extremes. every o1her ~ssibl e combination of the seven supply
options and three levels of DSM are tonSIdel"ed We used. MIDAS to evalullte
these plans under nominal conditions.
RI~ l"ugr4liOi'1 d Ruvl~
City of PIlW Alto J991iRP
42
6.2.2
DECSION TREE TO DETERMINE ALTER."iATE PLANS
figure 6.2
Rankin&: the ruuJts DC Ihe all-ccmbina:tioD Mnalysis
We 'U$ed system ave.-age rate and IOta! rescurce ccst to ranl the pla."ls Since data
for environmental externalities was not available 10 us at Ule time, we used lotal
resource cost inst~ad of societal cost as the criterion. Using these tv.·o criteria,. we
constructed two result tables. One contained a list of plans ranked according to
TRC and the other conta.in~d the s.:ame list or p[anss ranked a·:C'ording to SAR.
These tables are in Appendix E.
The first step in pruning the vast number of options WlS to select only those
common plans r..'tat were ranked in the top one hun&ed under SAR and TRC. The
rationale for doing this was that we were using TRC and SAR as rar.king criteria
and we wanted to choose plans that were favorable under both pe-rspectives. There
Ruow~~r /''It.jrlJlion &: RUJlf..,
o~ ofPak; Afro }99}!RP
43
... .:. .
,.'"
.... ...
",. ~. t
~~ ~~ . -".:: ;~
~ g'
~
~ ..
~~
<
"
Tabl.6,2
ROW COMMON RAl'lK • PLAN.
,'RC I ,,'AN
I 21< 6J I
~ )\. 10 2
1 20' 9J ,
4 ,.2 .,. ,
l 2" .. " " ,,>I ., 1
7 '" "' • , )01 " 9
9 1II , 1)
I. m • " II V' a " " lei , " " ~99 9 R4
" 212 11 .s
"I 20J " '6
16 216 " " 11 lOJ I. " " 227 " ill)
"
lJ) " ')() ,. 209 " 91
" 11' " " " ))2 )J ..
l3 J2J .0 %
" 196 " " " 10) J' 9?
([0 NOTIIINC
~ND·rOINTS OCCURING UNDER TOTAL RESOURCE COST RANKING
I< SVSTEM ,\VERAGE IlA'm RANKING -
SUP'LV RESOURCRS DSM LZVl.UZED
LEVEL R.\Tp;
SUfi I tiCI. I IJ{""AS SCtAN J.'IlF.NCIJ I /UlttS1.-Y f'CfiH ,..i(h/ll","
0 12 " • II " II 1.1 10.SI1
0 , , " 0 • • J.I 70J22
0 12 6 • 0 1 • 1.1 70.6-t6 • " 6 " ) • 0 1.1 I 70.646 • 6 " I, :< 0 • 1.1 'HHSl
II " I, " n , " 1.1 10.6B
II 12 (, " " " · 1.1 10.656
II " , " " 0 2 1.1 10,M, • r, " " 0 0 • [~ 1l.4 \1 • 11 6 • II 0 " J.2 1LoU6 • 12 " II 0 • 0 L2 'J1."st. • " " " • • " L2 71.495
0 6 • " ) 0 • L2 11.504 • 11 • " II • 2 1.2 71.517
0 " " 6 J 0 0 1.1 71.~21 • " 6 0 " • 2 Li 1J.5J4
0 " • 6 0 J 0 l.l 71.:)}4
0 I, " II J 0 0 1.2 11.539
0 12 6 0 II J 0 1.2 7L$~8
II 12 (, 6 • ) 0 1.1 71.5~6 . • " 6 0 0 • " 1.1 71.568 • 6 • 0 0 " 1 1.2 71.577
0 6 (, 0 1 0 0 1.1 11.603
0 • 6 6 .1 0 , 1.2 11,622
0 • 6 " • , , " i 71,628
10)S
Nom: ONLy TIm 'fOr lOOl~ND·I'OTNTS UNDER r.ACII11ANI(lNG CRrmll1A WAS CON,<;IDElmD
me
In,
U7.7
~J7.1
4)').1
"J9.1
08.4
04)8.6 I
4)R 8
4111.2
4)).5
04.0
Inu
04.6
4)4.'&
4j5.1
4)5.4
4H.5
414,9
4)5.9
'Il16,O
4lB
416. r
4)6.0
436.7
"36.2
41r..)
mil
•
"
,
,
, Table 6.3
RESOURCE
STIG
SeL
BGA.S
SGAS
FRENCH
RAMSEY
FCELL
DSM
LEVEL 1
LEVEL 2
LEVEL 3
were 25 pla.'1s that were common in th.e top one hWldred These prans had TRC
rankings ranging from 2 10 95 and SAR rankinss from 1 to 99 Table 6,2 shows
tht 20S plans that were common to both rankLngs
Using these results, we cOIl!>"1ructerl frequency tables (T&bles. 6,:) and 6 4} wh~eh
show how f'r!!!quent eac:b of the low and high state of each supply resource and the
three DSM levels .appealed in the 25 ptans
SUPPLY RESOURCE FREQUENCY TA.BLE
LOW STA.TE HIGH STA.TE
MW LEVEL FREQUENCY Of MW l.EVEL FREQUENCY OF
OCCURENCE OCCURENCE
Q 25 5 0
6 IJ i2 ]21
0 0 6 2s1
c 1 6 18
0 18 3 7
0 19 3 6
0 111 2 Sj
DEMAND RESOURCE FREQUENCY TA.BLE
FREQUENCY OF OCCURENCE
8
17
0
Heuristic analysis to select five alternate plans
The fil1a.! step In th~s stage of determini!>.tic analysis lJtl.liz:.ed the informa.tion lO
Tables 6.2, 6.3 and 6.4 and our judgment to seleci five nominal plans, The
follc\l\oing discussion and series of tables exptain that process.
IUSd4Ol:~ TII"Z'vuiDft &: RUlllu
c(t)I 4{ P;U'l tWO /991 lRF
45
Tabl,6.5
Piau
A
B
C
D
E
Table 6 6
Plan
A
B
C
D
E
•
~ -. ,
Slop 1
STEP 1 IN D[TERl~IINlSTIC PL4.N SELECTION
STIG SeL BGAS SGAS FREr;CH RAMSEY FCELL DSM
0
0
0
0
5 6 6 0 0 0 L1
Examining Table 6.2, we notice that the STIG project assumes it low state of 0
MW in alJ 25 plans. Neverthless., since the STIG decision was one of the critical
issues be'hmd the analysis and keeping in mind that the benefits of the STIG project
under ;mcertain C<lnditions has not yet been considered. we decided that at least one
plan should contain the STIG. We reviewed the table showing the results of 384
plans under SAR and TRC {see Appendix D) and we chose a plan that inc1ud(i the
STIG and ranked near the lop from SAR IU1d TRC pe"P'ctiv.s (see Table 65).
SI.p Z
Examining Table 6.3, we find that the annual baseload gas resource appears in all
25 plans. This prompted us to lnclude the b.aseload gas resource in all five
:c.~mic:aI plans. This is shown in the Table 6.6 below.
STEP Z IN DETERMINISTiC PLAN SELECTION
STIG
o
SCL BGAS
6
SGAS FRENCH RAMSEY FCELL OSM
o ~
o 6
o 6
s 6 6 6 o o o LI
R.6OVrC~ btUgrllliOJl &: jhJul!,J
CjtyofP~lo AlIa 1992lRP
46
.--."
.. 1$.
Tablo 6.7
SI2P 3
The SCt exchange resource occ:urre-d marginally more freq'Uentl}' in its low state
(6"MW) than its high state 02 M\\') (fable 6 3). Therefore we decided to con~ider
two plans with SCt in its high state and three plans 'A-ith Set in its low state
Sin(;e Plan E had Set in ifS low state, the remaining (our plans would have Set
occurring rwice in the low state a..id No-ice in the high. stOlte. We arbitrarily
assigned SCt ;r, P!ans A a",d B 10 its high stale and SCt in Pla.'ls C and D to its
Jow state. This is shown in Table 6.'.
STEP J IN DETERJII[)IIISTIC PLAN SELECTION
Plan SnG SeL BGAS SGAS fRENCH RAMSEY FCELL DSM
A
B
C
0
E
0 12 6
0 12 6
0 6 6
0 6 6
~ 6 6 6 o o o
Stop 4
The summer only baseload gas resource, in its high state (TabJe 6.3), appears i...'l 18:
of tho 2 S pi ans. This implied tIl4l it should be in 4 of the final 5 plans_ Observing
that P!an E contained the summer gas resource, we had to distribute the remaining
three occurrences or this resource among Pll1'ls A, B, C and D. The summer gas
resource was in.sert~d into Plans C and 0 since they had a smaller to-:.a1 supply
capacity than Plans A IIlld B The flnal occurrente of the summer gas Te50l!I"Ce
could randomly be i&lgned to Plan A or B. sine-e, at this point, they were identicaL
The results of this step are sho\o\o1l i:r.. Tabte 6.8.
&¥Nn. III~iMritNI & Ru,.ILJ
City of Palo Allo /992 IR1'
47
LI
,,---
Table 6.8
PI""
A
B
C
.0
E
Tabl.6.9
PIaD
A
II
C
D
E
STEP 4 IN DETERMINISTIC PLAN SELECTION
STIG SCL BGAS SGAS FRENCH RAMSEY rCELL DSM
0 12 6 6
0 12 6 0
0 6 6 6
0 6 6 6
5 6 6 6 o o o
Step 5
The supply frequency Tabte 6,) shows that each cfthe bydro lesour~es. French and
R.amsey~ 'W'ill occur once in the five nominal plans E"amining Table 6.2,
However, we see that French and Ramsey Deyer occur together in any of the plans.
Therefo!'t, we detennined thai French wHI occur once and Ramsey will occur once
amon,~ Plans A, B. C or D with the constraint that they do nOI occur in the same
plan. We excluded Ptan A flom having .. hydro reso'W"t;e sLnce it had the highest
amount of ~upply at this stage.
W. "'er. left with assigning French and lUmsey among Plans B, C, and D.
Having no other rNSOn to choose among these plans other than to make sure they
did not occur in the same plan, we arbitrarily assignt.1 French Mudow 10 Plan D
and Ramsey to Plan C. This resulted in the following tabl •.
STEP 5 IN DETERMINISTIC PLAN SELECTION
L1
STIG SCL BGAS SGAS FRENCH RAMSEY FCELL DSM
0 12 6 6 0 0
0 12 6 0 0 0
0 6 6 6 0 3
0 6 6 6 3 0 LJ 5 6 6 6 0 0 0
IiIKIfU'C' JrtU,,-oIj(Jtlll .. R.Jufu
Cil'y of Paw AUt) } 9J)] lRP
48
• •
•
• ,
Table 6.10
PlOD
A
B
C
D
E
Slep 6
T~le 6,3 implies thai fue fuel cell should OCCI.lf ~n h .... o plans. Since the fuel cell
is a new technology and keeping Lt! mind thou uncertainties are yet to be
considered, we: detetmined that the fue! cell shOllld b~ included in. one plan onl:",
W"! selected Plan B :since it had the teasl supply capaci!y. This is shown in T~le
610.
STEP 6 IN DETERMINISTIC PLAN SELECTHlN
STIG SeL BGAS SGAS FRENCH RAMSEY FCELL
0 12 6 6 0 0 0
0 12 6 0 0 0 1
0 6 6 6 0 3 0
0 6 6 6 3 0 0
l 6 6 6 0 0 0
Step 7
Examining Tables 6.2 and 6.4, we see that level 3 DSM does no! oc<:ur in any of
the plans beLng considered. We kept in mind that the methodology ~d to
construct the frequency table, eq"a1 weighting to IRe and SAR, could have been
prejudiced against selecting level 3. dut 10 its rtlatively higher rate impiioct
Further, the way in which the DSM progr=s were designed cculd have given level
J an upward rate impact "'at cculd be mitigated by redesigning our DSM programs.
Therefore, we decided to add level 3 of DSM io Plan B, due to ils low supply
commitmt:nt The frequency Tablt 6.4 implies the:! Jevel 2 DSM should occur close
to three times in the five nominal plans. Smce Plan E contains levei 1 and Plan
B contains level 3, Pians A. C and D ~utomatica1ly ..vert assigned !evel 2 OSM
Table 6.l1 summari:z.es the completed nominal ptans.
FU~OW'N r""r-alion & Ruulu
City of Polo Al'Q J ~91lRP
49
Ll
...
\
Tabl. 6.11 STEP 7 IN DETERMINISTIC PLAN SELECTION
Plan STIG SCL BGAS seAS FRENCH RAMSEY ICELL PSM
A 0 12 6 5 0 0 0
B 0 12 6 0 0 0 2
C 0 6 6 6 0 ) 0
D 0 6 6 6 J C 0
E 5 ~ 6 6 0 0 0
6 . .3 Uaee-rtaiDty Analysil
The deterministic: analysis coupled with somt" judgment caHs, up 1:0 this point. helpi!d us
reduce the 314 alternative plans to 25 plans and finally to five plans. However, ability of
each of these plans to minimize SAR and SC is subject 10 uncer..ainty. The prior stage of
analysis assumtd nom~nal conditions for different lLnce"ain events. This. stage of the
analysls was performed 10 analyze how each of the above plans \oVCuld behave U1lder
varying assumptions or an uncer"'..a.in future.
~~e started with. identifying uncertainties we belJeved would affect the: different plans.
Ntxt, the sensitivity of the 'Wlcertainties was measured by toggting each uncertain ¥ariabte
between its high and low states and obS-'rving how the change would affect the value
criteria.. This process Iimitrd the number of uncertainties to those most releyant to our
futuce plan af action. To gather detailed information n:sardlng the probabillty dtstrihurion
of the uncertain variables, we J"~lied on a combination of stroJcnU'~d interviews of experts
add examinations of historical data That done, we were in a position to analyze the
effects of uncertain conditions 00 the five piws. After running the m~~l. we calculated
the expected value. e,.;,treme values and c.umula~ve probability distributions for each plan,
and these talcul~ons led to our tKOmmenoation.
63.1 Identify,nz the unc:ertain \lari.b~e$
IlffluDlU diagram
In order to identify the uncertain variables ar.d focus our discussLon, we decided
I.e structure the analysL!Ii llsing .an influence diagram, which is " grsphlc..al
representati\Jn of the declsion problem. We used a compl.!tcr model called Deeis~on
Programming Language (DPL) to he[p us draw the influence diagram. The
resuJting graphical representation of the problem h~lped focus the discussion from
a fuzzy gen~raIization to a clear undei':!lo1anding of wha! each p~rson was attempting
RuOJU&" I1f/.",aliOft &-&S1IJr:
City of Palo .<110 1991 lRP
50
Ll
LJ
Ll
Ll
LI
'; .
to communicate.
We conducted a few group meetings 10 c!a.~iY the influence ciagram, ending up
with the on~ sho'olr"'ll in Figure 6,)
ri'nUENCE D1AGRA.~llDENTm1NG
l'NCERT AlNTIES & VALUE CRITERIA
~
I pi"
FigtJre 6.3
The wlcertainties that were identified are:
1) Forecast~d load
This refers to 'ihe forecast of electricity demand and energy over the
anaIysjs time hori~n of 20 years A detailed description of forecast
R'I4ru'e, l"uzraJum &: Rnufu
City of Pclo ""/10 199J !RP
51
methodology is con.ain'.!d in Cbapter:2 Toe high, nom.inal ar.d low load
fOTecasts were ::he uncertai."l. outcomes
2) Hydro I~.l
Th~s unco:rt~nty was rele .... ant to the de..:ision due to our exlsting .esource
mix and our possible future resource mix, The output of the CaI'lveras
hydro project,. and itS future enhancements, is direc:tIy dependent Of! the
level of ".vater in itS rese,votrs. The best case of thlS uncertainty was
mode!ed with data repr~nting an a'w'erage wet year, atld the worst case
~d data representing an average dry )' ear
3) We$tenr ene'KY dllocation. in 2004
The existing contract with Western links our energy allocation to the system
loa,j factor. There are proposa1s from Western to reduce customer energy
allocation when the wntract has to be renewed tn 2004. We assumed that
the b-est C'lse ot this uncertain event was maintaJning Qur current energy
allocation. The nominal case would entail a 5% reduction and the worst
ca.c;e would ental! .. 10% recluctivn in energy altocation from aur present
entitlement.
4) DSM penetrollon
This uncertain variabl e refers 10 th e participab.()n 1 evels expected for ea.cb
DSM program.. This variable W"~ included because it is very difficult to
predict customer response to DSM programs.. We modeled this uncertainty
with high and low penetrztion levets.
~) DSM imp4ct
This variable refers to the energy and dernmtd reduction on the system load
due to the implementation of DSM programs. Most of our DSM entails
replacing an existing piece of equipment "With a more efficient one. To
forecast the load reduction correa1y. we had to know the energy tne of both
the eqwpment that would be replaced and the efficient equipment Since
the DSM program desjgn is within our control, we ha .... e a 800d idea of the
energy use of the efficient device we 'Wish 10 install. BUI, because we lack
perfect informatior.. on all existing equipment, there is uncertainty in
estimatin8 energy use of the old devic.es. This uncenainty was modeled
using a high and low impact.
iWlCW'e_ bf/~~tiOft &-Ru-.ltl
CiIY of Pow Ai", J992 /RP
~2
6.3.2
6)
7)
, ..... -." ,-~.,,>. ,
" ,
Our existin8 contract .... 1m Western allo~ u.s 175 MW of c2pacity. Since
WAPA !.U already proposed 2%, 50,."" and 10% reductions in capacity
31Jocation for the post-2004 period, it is highly probable that t.'1e contract
wit! not be renewed a! its existing capacity level. Therefore our low,
nominal and high case capacity a!hx:ations 2!SSumed redu(:rions of' lO~/6, 5%
and 2%, respectively.
Natura! gtU priu
Since many possible future resource aCQi.Lisitions are natwal gas fueled, and
keeping in mind the volatile n.2lUre of the gas market, we included natural
gas pt"~c-e as an t.lr.1certaln vuiahle. The data regarding high and low natural
g~ prices were obtain~d from esc.
8) Rnource llWlilabi/ity IUIII cost
Resow-::-t::" availabllity refers expticitly to the fuel cl"!lI. The fuel cell
tedl..'1oiogy we are considering is the molten ca.rOOnate fuel cell. The
demonstration or the lechnotogy is scheduled for 1995 The commercial
production of this futl cell is scheduled for year 2000. The contract 'With
the fuel-<:ell manufecturer specifies that any cost overrun will be absorbed
by th~ manufacturer, protecting the Ciry from zost uncert.ainty on this
resource. But since this resource is on the cutting edge ClfteC'hr.ology, th~re
is a definite possibility that the manufacturer may not meet pr..,jecled cost
and deadli1'\e, with the result that the resource will Dot be available at aIL
The cost uncertainty refers 10 the baseload gas resource in particular.
NCPA had not decided at the time we were doing the analysis 00 which of
several baseload 8as resources to pursue. Since the price of these r~u1Ces
are different, we assumed the cost of lite b~load gas resoun;e as an.
uncenainty.
Determininl the stnsitive uocertain "anables
The list of Ullcertainties -ilia' had been identified was shortened for two reasons.
The first is that we wanted to hmit our analysls to variabtes having the greatest
impact OD the de,.-;.isloD. Th., secol1d is that the modeL we use, MIDAS, can
evaluate a finite set of endpoints which is lim,ted by the amount of conventional
memory available on the computer. This constraint limited us to four uncertainties
with each uncertainty b"ing allowed to assume three !>tates (example. fore-c&S1 Io.td
could be high, nominal or low). The decisioc tree struC1U1e used to model this part
R.8o.r.:tI 11f1,Pfion dr RUllrU
City of Pale ..(1'0 1992 IRP
53
'. -~
of the analysi'!; is sho~ll in Figure 6.4
DECISION TREE TO DETERMINE SENSITIVE UNCERTAD.'TIES
!.~!
PI.u; A Nominal State
PIa.."lB
r---
LowSt:de
His'! SU_" __
Nomi. ..... 1 State
/ L.ow Stale
High State
Figure 6.4
<J
<J
Each wH:ertaln variable is evaluated or.e at a time. Though ~me of the uncertain
variables could be correlated, we did ll~t assume any correlation between the
uncertainties. The decision tree was simulated for each uncenain variable and Ll-je
SC and SAR were evaluated for ~a.::h pia.i. The results are graphed in Figur~s 6.5
through 6.12.
Rt:~rr;:t In/_grafion &; Rtl.wlu
City DJ P::Io Allo 1991 JRP
,
54
•
,
SENSITMTY TO l'ORECAST LOAD
TOl:ll Rtsource Cost -I
... 5::~ I '
f '" 1_ -_~~-.., [I
j"'i -':-E~"'III , "" L ==;;::==+==,=_..c="~:::_ i r· --::.----: ~ "'" "" E 42"r=--_ A ~ I I
,. ~w._1 A--~. "-C 1'\an D "-t: j
r~==================·~l
Socitta I cost
,--
I System average r2t~
i "[-
I ~74---I~"~ ;"-..
-7!:p ~ ............ -= • • I
in~
l ;'~p -.-o-.--"..--C--.-,,-o-O--.... -.
RumJ.rc~ rntrrotiim &: Rtlldlt
ell)' of Palt) AJ10 J99] IRP
-', ,
Figure 6.S
I
E~I l~~
~--
j -H., .. :.-; i !-C-B1M '-.:i :1
I-Lo .... J
i
J
55
j .,>'
SENSITIVITY TO HYDRO LEVEL r--------'~
II Total Resource Cost I
I ~'" I i 'timE I .1 · '~"'I I
II ~:: -----------::;:-----..-----II =: I i
i oQ.St~" i ----=:::::::: -----', I 1,-::l::':~:':~A==-"";:-:'~~~~""~~;c~~~~"",~~-o;=-:-",;;,;,;;;~===::= ___ J '--I
Sodet.:ll cost
.... ,
System averagf rate
73rl---
i 74 \-1-----
i ,,~
t "L=::::::=-:, ---J 71~::--."::
70L •
RU(JllTc, I,.,16pariOfl: &: &31I{u
City r;fP4loAllo 199] lRP
Figuro:.6.6
......
, .
I r-; wei II
bJI
I
J
1
i
i_~·~I[ I~~ .. A.' .. ·II j-_o:J
56
•
•
SENSIT~TfY TO WAPA ENERGY IN 10Q4
r
Total Rr.sourct C6St
i:r-I ---.i ." L ____ _
!.w\. ~ I ~!"'I=---~ -. . 4" '---------
Ph"A PbnB n-c ~D PWI!
I
I m SQci __ e_tll_I_C_o'_t ___ _
-I 11:bs-;-:Z= .~ 1'''(---
i-H.i-l~~;'''''I..'';1
\ -Cr--D,·'W'.cv J
i _,_~~
r~Hi.ll)(.','.arl..l" \
I = :::::~ I
! ~A~. Pbn_C __ """ __ " __ ..... _' ___________ _
I SY5tem average rate
I
I
7~ --,,_I __ l n L. -C>-____ _
; ". I _
1"~ -, n_
_ "' --=:::0-, I
1~L--
.... ,n
Figure 6.'
R'.1QIUCI 1"r.r,jM &-RUlllu
City 0/ Palo Alto }992 lRP
/-~ -l;~;~-;-I
1-=-:e,·9.s~-.a:tF ~
L=--u-~~U' I
•
5i
SENSITIVITY TO DSM PENETRA nON
System A"erage R.te
~ :: t-----C ----~ n f-~-----------L------""--~--! "t== : :: 5,,_--~
" .--------------
PI ....
IWSIIfII"« [rr'.yollcm & llc.n.du
City of PolIo Alro 1991 IR.P
Flua Plo ••
Figure 6.8
'--1
I
I -',
1 -Hi,. "''' I
I --::--.,.""
I , I . __ -"'" ,e", t
58
'. ;' '-," . ~'.
SEl'."SlT!\'ITY TO WAPA ALLOCATION IN 2004 r-. Tob1 Res<>urcc Cost
---------,
I i:C ---
1 ~1~ ~------------
.(~i ~ i I ~ I 4~~-=~~'~~~~~~~::::----
t <00 ,'--________ _
Plu! A pw. B J'1ua C Pta. D Pbn E
R.$()W'~t brlerr13ti07l &-Rtn;IIJ
err) of PIlIQ Allo 1992 JRP
Socieb1 tOS t
System a\'erage rate
Figure 6.9
1~-,,· .. ··1
( -:;--s •. '~·'.I
L _______ ~·~·,
E -
-Ho..!t!~~
--=-lh·9~·" l=---l.o. }ow. I
I
I
•
59
SENSITIVITY TO NATURAL GAS PR~I~C~E,--_ I Totar Resource COS! 1
'
I . '''L I
I
i 1:~1 .-I=~~':ll
I .:...._-+ __ .... ==-"'---~ I ~ "''''po i i, l. (l" i' I -----
I' "" ~I ---Ii
. p~ " .Pt&n • P'.-": PI.III' D I'I..uII E
Societal tost I i : rl----:_-:_-___ _
i::t;~_
1 : l-i ---'------------
, lWIA P\u I ...... _ .
L--__________________ _
I SysteI:l average rate
1 73,.-,-----i .,,}--
I ~"t--I -{,.~--
-1l I
"L. ____ _
R~~. Int~grori(Nl .: .--:..esult.z
City qfPo'oAlto J99} lRP
Figure f;.IO
.<
-----l<Iwp.$
r-------
i -Low.., II
---.:;.--&so, pi I~ ... , ...
60
SENSITn'lTY TO DSM IMPACT
Total R .. ourct Cost
--~~-----
E·"~l -~""'_I i I ~L4""" i i
~ __ J
System a,,'erage rate
"-i-----
4 7' I i",-----
) : ~~-~~9~_~i-~=---~ __ ~~~~=
11IIn " pj."n B PU .. C P'.u. P Plan. II
R.~0tI1"C. JIIU".02timl &: Ru~l/J
Cily ofPtI'o .1llo /991 IRP
Figure 6.11
61
SENSITnTfY TO ANNUAL BASF,LOAD GAS COST &
FUEL CELL AVAILABILITY
Tot31 Re!5C'urcc Cost
Societal cost
System anrage rate
!~t7----.-
J ll~t . ~--~
" -----
PtanA p~" B PluoC ~D pt:,.1lI.
h30fD"U II1tllgnttiOlJ &: &n;/lJ
Cit;l of Pal" AIt.3 1991iRP
~ .. -. ,
-~.\-r;::1
-. =-9G .. ..s. -.<C I
--~Q ...
-I
I
E .. -.= :::lll t=--~c?A I
I
J
62
•
•
•
•
Tabl.6.12
Examining tilese resu!!S, we concluded th<:t the lvad forecast, hydro level. Westem
energy allocation in 2004 lUld DSM penetration unt:ertainties deserved fTlore
attention d"IJe to their reliUive hl&h~'" impact on SAR and socletal cost and we-re,
th~refore. selected as the sensitive variabJ.es. Table 6.12 shows the lis.t of
unce:uinties that were initially ide~tifjed and the uncertainties that were selected
for further walysis.
UNCERT.U'; VARIABLES 8. SENSITIVE UNCERTAINTIES
FNCERTAINTIES CONSIDERED UNCERTAINTiES CONSIDER I'D
FOR DETERMINISTIC ANALYSIS FOR PROB.~BILlSTlC ANALYSIS
Fore<ast Load ForKast Load
Hydro Level Hydro Level I
Western Energy Allocation in 2004 Western E~ergy Allccaticn in 2004
DSM Perietrztioo DSM Penetration
Western Capacity Alloc.ation in 2004
Natural Gas Pri c.
DSM Impact
I Resource Availability &: Cost
l.ooking at tlu.se plots. we observed th, copflict betWeen oW" ChQLCe cor value
criterion. For example. examine Figure 6.8, the sensitivity to DSM pe.'1etrarion.
The plot ofsocieta! cost implies that achieving nigh participation in DSM programs
would !ead to lewes! societal ccst, irrespecti .... e of the ptan we follow. On the ether
hand,. the System Average Rate plot implies th-at a tow participarioil in DSM
programs would result in the lOWes1 levelize.d rates, irrespective of t. ..... e plan we
chOMe ODe has to make a trade-off between so,i«:taJ cost and the rate impact
This mdt·c:ff is Q.oi'le implicitly u-. our analysis as W~ i;ooose one plan over the
othe:r. FUMe' research LrItl) thlS area. was warranted.
6.4 Ass.e"inC prcb.:tbilWes or uD,emin .... ariablrJ
Once the shortlist of uncertain ... ·ariables had been determined, we needed to assess
probability distributions for each variable. There are two schools of tho-ught for
IUJOW"CI II1UP'dlian If: RuullJ
Cl1}l of P<J/o _~i." }992 iRP
63
quantifying uncertainty, objecli....-e and subjective. The objective view characterizes
uncertainty as (I) a. property of the physical "IoVOrld. {2) I. statistic based on repeated
e1C'periments and (3) only making sense for repetitive pbenomena:. The subjective view
cna13cteri.zes unc-ertainty a.:; 0) & belief about the real world, (2) dependent on an
individual's t,ehefs based on current lnfcrrnaticn and. (3) muing sense ror any type of
phenomenal.
Each view of uncertainties has its O\lwTi appeal. The objective school of thought wants 10
provide unbiased ¢pinions and also believe5: that if & phenomena has never or rarely b~en
observed, there is no way of estimating its hkehhood of o.:curtencc. The subje.::tive view,
on the other hand, is driven by the !'lecessity to make decisions even if there is no evidence
Gor prior data on the behavior of & phenomena.. Tne sl.\bjective v~ew of assessin&
probabilities was foremast in OUT minds because we bad to decide on several issues and
we did not ~ave enough time to edlaustively researclt each utlc~rta.inty
In order to assess s~bjettive probabihties. we interviewed experts using a process
structUred to eliminate biai. In addition, we examined 'historical datz. for some
uncertainties and determined the probability of occurrence of each state.
Within thi.!o fra."Tle'Work, we as.oc:essed the probabilities of occurrence of each state of ea.c:h
shortHsted uncenalr:!ty. The results are shown in Tabte 6.13.
Table 6.13 PROBAIHLITIES OF SHORTLISTED UNCERTAIN VARIABLES
Uncertain "'ariable Probability
Rilh Bas .. Lo",
Forecast load 0.25 0.50 0.25
Hydro level (wer,.ave,dry) OJl O.H 0.11
WAl'A ent:rgy allocation ir. 2004 0)) 0)) 0.33 I
DSM ]>enetration 0.25 050 o~
2Set "-Cuaf'i'ifyinll JIMdpeMtlll [i~CU!ajMty: .\lIrhodM-ogy, bptlril!.f'icO, dM llQillhu: by Miley W.
Merkhofer, IEEE TranNctioJ'\S on S),SU;tnI, Man, alld Cybernc:tics, VoL SMC.I1. No . .s, ~ptlOct. 1987, for a
dettWcd npoaiti(ll'lo cn collducting such inttrncw$ .nd tht pros and tons of NCb an .pptoa,;h.
Jksowrc, 1""Z'4ffOft &-R,.whJ
Cil)' 0/ Paw .41tQ J'l91 lRP
6.S Coostruelina the-decillIon trte
No\'", we ~ould construct the decision u-ee thiit would represent the proolem we had
formulated at the outsef of the analysis, (See Section 1.3.2. J for statement of the problem.)
flgurt 6.13 shows the deci~ioll tree that incorporated the five alternate plan~ ar,d tlle
shortlisted uncertain • .. ariables. This decislOr.. tree has 4C~ endpoints. Each I:'ndpoint is a
Wlique representation of a po:iSible future scenario. TI\e de,isioll node had five decisior::.
bran,hes. Ea~h deci~lcn briU1ch rep!ese."ncd .. -..n.iqut" 'Plan thM ~ontiUned ~ ccmbil".a.tion
of supply and DSM resources. Each of these aIlemative plan:> could be aff~cted by a
variety of uncenainties. The previous phas.e of analysis had shortlisted those uncertainties
we believed to be relevant to the decision .at hand These uncertainties were DSM
penetration, W.o\PA energy allocation in 2004, bydro level and forecast load, Attac~ed to
each decision branch was a series of un,ertainty nodc:s, tach node having thr~e uncertain
outcome branches \oIooith their associ~red prob.abihties, Tra.versing the upper branGhes of
each node yields a plan and associated future described as Plan A, with high DSM
penetration,. high W AP A energy allocation ic. 2004, average wet hydro years and high load
On-"! can see the .... ariety of futures analyzed for each of the plans as we progress through
different paths of the tree.
AI this time, we al~ re~va1uated the SeL contratt to determine if it was appropriate to
resize It in our artalysis. We yerformed .. determinlsric analysis In whic.h e.ach of the p!ans
identified in Tabl, 6.11 w.re tested with the SCL <i .. v.ned between 6 and 12 MW. 1be
rtsUlts of this analy£\,S: prompted us 10 resize the upper and lower hml15 "r this C'Ontract 10
II MW and 9 MW. rospe<:tively. Tabl. 6.14 ,hows the "vise<! plans with the moo,r,ed
size On the SCL ct'ntract. The !otDDAS model was set up 'Wit"li this tree structure ~figure
6.13) and each of the 4;05 endpoints was simulated over the-analysis period of 20 years.
Table 6.14 FIVE DIFFERENT PLANS USED FOR ANALYSIS
(MW)
STIG SeL
PLAN It. 0 II
P1.A.''18 0 II
PLANC 0 9
PLAN D 0 9
PLAN E S 9
RrSOU1'er I'''_zratum &-Ru",lu
City a[P.::.lb ."-lro 199] IRP
BGAS
6
6
6
6
6
SGAS FRENCH RAMSEY FCELL DSl\I
6 0 0 0 L2 -30
0 0 0 2 L3 -33
6 0 3 0 L2 -30
6 3 0 0 L2 -30
6 0 C 0 L1 -20
DECISION TIlLE FOR PROBABILISTIC PlhSE
A
DSM
penetration
Figure 6.13
6.6 Results IlId Recommendations
Fore_g_g
load
The-tes-~ts of the analysis d~rited ~n 'the previous. sectlcn art ptesented using cumulative
probabllity distribution curves and expected value of each decision a.; sho,"",'!] in figures
6.14 and 6.1 5, Funher~ these results are piest:nled for both our value (:riterion, System
Average Rate and Societal Cosl
Expected '\'a1ue is t.'1e probability weighted average of the outcome of each decision. The
best decision would have tht lowest expected value. In addition to th~ expected value we
ace also intetes1ed in the extrelne value of -each decision. The differen,f'! betvw"e-en eX-seme
values of each ptan is a. meSSUl'e-of the risk associated Y<1th that p1an,
ElUmining figure 6.15 we conclude the follo\.\r"i.ng. Plan B has the lowest ex.pect~d vah.:e
of SC !Ild the highest expected value of SAR Pian A has the loweS1 expected YI1lue of
SAR Uncle! a WOT5t case scenario, plan B has the highest SAR impact Under a wont
case s.c:enano. plan E has the highest societaJ cost impact Examining figure 6.14 we see
that the plan with we bghest probabLlity of coming W'lder any specified societal cost value
is plan B and the plan with the lowest probability of comif.g in und.er any specified
societal cost is ptart E, Looi:.lng at the plOts f.;}t" SAR we see that p[an B has the. t'lwest
probability Cor achieving a SAR lower than a s"ecified value. The other plans have
approximately the same probability of achieving a SAR lower than a specified value. From
}\UDiII"C. ll1ugratifm de lU-.twlts
City of Fol<; Alta 1991 lRP
•
•
•
,
•
•
. " .. -:
CUMULATIVE PROBABILITY DISTRIBVTIOl\'
--------,
SOCIETAL COST
I
--! 1-·-·~!.l.:lA
!
j
1,_-p.1aca
I-"'~' G;;:~
I
I I ... "~A
I
j----=--F~B
I-~~c
I-Fl.u.D
---..,..-P:.v.:e:
'--
:: ;'-.-. -~-'. -~-. -~.-g-. -~-~.-~-. -S!-~-$-. -';l-~ " i :; • ~ -I
~ ~ e ~ ¢ Q Q 0 ~ 0 = e 0 ~ e C Q 0 ~
Cumulative probability ------
Rt.to",,.r:r ],.I,gralicm &: RUliit...
Ci:y of P"/c Alro 19f)} lRP
Figure 6.14
___ J
•
\.
67
EXPECTED AND EXTREME VALUES
SOCIETAL COST
.." I---~
51S
'; '50 I rr "'"
,m f'"
11: I
i----~ -
::: 41S. /.' u k I. ~50
"S ." " '.
~:r
B C D E L_A __ Plon ~----------------~
SYSTEM AVERAGE RATE
r-----:::-___ -----------
r
~"
70 f--I~ .. ~.-----.-·-1_",",.---1 ... =.---·-.. ~----------
!te»llTel In.urrariOl'l cI llul)iu
Cil)r ofPQ/co Alro 19~] l1U'
B C
PI",
D
Figure 6. 15
68
•
•
•
Tabl.6.IS RECOMMENDATIONS
SHORT-TER:vI ACTION PlAN
· Parridpat~ in the Seattle City Light Energy Exchange
· Do not participate ir,. the NCPA STIG proj.eci
· Implement pilot DSM programs with DSM level 2 (:30 MW) as the long-term
wg ..
LONG-TER.'I1 FLEXIBU PLA:'I
· Participate in stlJd)/ of Calav.ras (French &; R&msey) hydro enhancements
· Participate in study t;lf annual baseload gas projett and summer-only baseload
gas project
· Pursue 75% of DSM achievable potential (30 MW)
· Investigate locaJ generation and co-generation options
R,MJfI.rC' !"uiT"lion d: lUwlu
ell)' of Pow Ailo J 992 lRP
I
I
70
,-",,-
.. i!
•
•
•
•
•
•
•
APPENDIX A
i. •.
LESSONS LEARNED & IDEAS FOR
IMPROVING THE 1994 IRP PROCESS
, .
As Winston C~urchill :said, "The plan is ncthlng, planning is :verything' Once we completed
our analysis .. .,d !U.rt-ed to document OU1' results. we could ioel: back and critique the 19092 IRP
proct$s. This is L'"l integral pan of the continuously evolving phuming proc:ess. This sec:tion
Qutiines ~e feedback-loop we us~d 1.0 iron out t."e wrinkles we encou..,tered white de ..... eloping. the
plan. Sta..~ meetings produced structured criticism, enumerating an exhsusti\'c list of 9J:eas
needing impro ..... ement These criticisms can. be grouped into four areas: pra<:ess, data., time lind
commlJ.D..icabon. Suggestions to mitigate criticisms in one area imp.lCt other areas. For e~ample,
improving data *"uracy could lead to I streamlined process and thereby contribute to completing
the anaIysl!:1 in less time. To ILvoid repeating our 5t.1ssestions in each subjet:rive area, W~ -will let
our 5uggesti0l'2S and criticisms fan into one relevant af'=a, l<eeping in mind the s)'nergy inhere!!t
in many sugg6stlons.
Al Dat1l
In developing the data sets rOf the varia us compL'· ~ ~ model S, we noted a recurring thtme
of data inCorlslstenq. Data was 8athered fror numerous soun:es &nd each. model
accepled data in a parncuJ at" format and with a certain level of ac curacy. Also, the
vintage of some data was questionabte. To mitigate this flaw~ we intend to creatt: a
computerized common data ""t-,ase and centralize the rtsponsibihty for updating and
maintaining that data. The data ~asa wiU provide load fortea-ct, supply resource and DSM
mUSUrts retated inf'ormatioc<
J..{uc:b of the d::tta used in the DSM analysis originated from the SMUI)/XEN"'ERGY data:
base and staffs empirical e:cpericnc:e. Tht:re is a conseMllS that we must continuously
update the data set for DSM measure Ioadshapes. This is essen.tJaJ not only for planning
purpoSesJ but also fOf evaluation and verification. To irr.prove: the accuracy and relevancy
of the loadshapes we will \o\IOrk and/or interface with CEC, EPRl and outside consult.ants
to ensure that the data we C('Iltect is ;~levant lo Pal" Alto. We also intend to leverase
PG&E rese"3Teh iT. to DSM loarlshapes
Finally. we rely on outside age-ncies such as NCPA and TANC for managing aur
resources and performing some of our planning activities (such as NCPA's RFP prcc~).
This me3.""lS that we ~ouId be more assertive in acquiring the Clost updated informatioo
from these a~ef!cies for inclusiIJrl. in the data base.
A.2 Time
Deci:.ion analysis hetps us spht the complex rRP process into ma.nageable &nd intuitive
steps, b"i.lt we ~ar:.not take ad .... anta.ge -of this excellent framework. . without a.Iio-cating
sufficient time to perform between-stage analysis. We felt there was not enough time to
U.uc.ru ~4"u:d and id<o: fur imprO\ling Iil~ } 994 1RP proaSJ
Cil)' ",Palo A'!.; 1991/RP A-I
•
properly judge th.e results of uch step, and tf:iat the entire d~~ision, IT';L.;.ing process
su..'Ter.:d as a result There was a una..,imous feeling that adequate time may have bdped
to streamline 81ld boost the quality of t.J,.e process. Howe",·er. we should note that
experien~ gained from t.iis and future IRPs should reciu':e the need for more time.
'Nt need to c:'l;sure a.dequate time for major o!' mlncr [evisio~. if any, "Hithout affecTI!'og
the final rime Im~ . .Fu.ture planning efforts will incorporate schedules. Qu~tifylng
~~n·bours is required to complete e.;,h task ot the IRP. Our experience Vrr1th this IRP
is I good Ii rsr S1ep in doi 08 that
We learned from our first e~ert interviews that more time and p!'a~tiee were needed to
obtain ac,urate and credi~te results. 'Ve.eJso need to iden~ify more than one expert and
give: them the background necessary to answer questions specific to Palo Alto
Utilities across the country &re developing IRP's, We intend to monitor their progress and
incorporate their advancements to avoid reinventing the wheel. leaming from the wealth
of informatioa available Llv"ough PG&E. the la:gest utility in tho! country. we could
significantly reduce-the time required: for DSM planning, evaluation and implementation.
It is in our best interest to develop and ml." .ta.in contactS with PG&E in order Ie save
time by sharing data and tecful.iques. We should also pursue this strategy with
Sacramento Municipal Utility District (SMUD). l!I\other visionary utility in th~ area ·.,f
DSM planning and implementation.
A.3 Pro, ...
Senior man.a.gement input at the oLitset and at critical junctures of tne IRP process serves
as a guide to analysts. Using CATALYST, a structured pro<:ess developed by EPRI, will
beJp to unite the broad, qua.h!a.tive aspect of senior management with the detzi1ed
quantitative perspective of utility planners and analysts in a practical and timely way.
The goal of CATALYST is an analysIs that both planners L"l.d executives see as useful
and c:ompt:!hensive. We intend to use this technique, or an equivalent one. :0 develop
futurts :111d asso<:iated strategii!s to help streamtine the reso .. rce screenirtg process,
CATALYST can also be use-d to develop joint/conditional probabilities that quantify
dependent unc:trtain \'ariab!es. As an alternati .... e to CATAL YST~ we could consider using
snake diagram proc.ess 10 schedule inten.ction between reviewe1s and analysis team. W-e
could use computer models like DPL to aid brainstonn:ng processes. Making a list of
what is desired from upper management and ineiuding spe-;:;ific tasks and the number of
hours needed 10 comple~e these tasks would en-S\.lfe that upper management allocates
sutTlc Lent time to th is process.
Since the value criteria we used (societal cost and system &ve:-age rate) led to conflicdng
resource choices. we need to clarify the decision makers' trade-off between the criteria.
We should study the pros and' ccns of detemlining a singje objecti .... e function (OF) v.ith
USJOf'U 1,(U7wl and id.!o! for ,,,,proving the 1994 IRP prQ';:US
City oj Palo All" 1992 lRP A·2
..
• ,
multiple criteria that encod~ de.cision makers' jlJ.dgm~nt and trade-offs_ We should test
the set or plans ',gainst difftrent OF's, each OF being representative of a certain viLlue
We could atso test the set of plans agai,1st different 5C'enarlOS For example. we could
ask, -Which plans do well under ",averse hydro conc.itions?" Wt could a!sc use objective
fwtction formulation to balance the rate impacts, of diff~ren.1 resource choices, ovl'!! th~
first 10 yean and seccnd iO yeus of the: 3Ilalysis period.
Some DSM programs and suppl)' resources. have lifetimes that do not match the
spending/e"Pense patterns a.ssociatt!d with the resource. For t)Campie, hydro projects are
typica!ly licensed, and therefor.:: accrue b(!oefits. for SO years. Meanwhile, !hey are debt
financed, or costed (lL:t. over 30 ye1Ts. Th~r~fo[e, the COsts are fully counted whereas the
benefits are only partially cOW1ted~ These residual benefLts that are not counted make
some supply projects i!.!Jd DSM programs appe.u-over-priced. This is an artifa.ct of Ollr
planning method ar.~ .... 'la.!ysis horizon tha.t can be a.ddressed by asS\lmlng that t..i.ese
resources or DSM pr )grams. are debt fmam:ed ov~r its life cycle to synch:oniz,c: th.e stream
or expenses and b~n~fits .
The above sugge.mon will ntcessitate a larger data base. Rate impacts using levehzed
data set may not reflect the actual rate patterns produced by a lumpy investment like a
DSM progymt cnaracteriud by capital expend.itlJres in the initial years of program
implementation. LcveH.rit.g the costs over the lire cycle of the d-:vite, wr.ether it is
supply-Or demu:.d-side~ 'Will ensure that COsts .!.!"d benefits associated lNith a resoW"cc are
considered equitably o .. ·er the analysis period. Follo'Ning this suggestion would entail
preparing two data sets, leveliud (or debt financed) cosu to do the TRC tracking and
actual costs to trade rate impa-;ts.
Life cycle .and degradation of DSM and s.uppiy TtSOUTCes is an area that des-er.·es
attention. On the supply side, .. particular example is the. fuet celt stack rtptac.ement
cycle. Ongoing research and results (rom the Santa Clan.. demonstration pr.oject will help
10 obtain mere insight into this problem. for DSM technologies, life cycle refers to th~
manufacturer's namep:.ale rating. Degradation r:fers to the remo ... ?>..! of a device (eg buJb)
before the end ofits usef, .. l life and the pr.emature fai!ure of devlce before the nameplate
bfetime. Presently, degradation of devices js addressed by reducing the hfe c/cle. W-e
will study the possibility of explicitly accounting for p~rfonnance de-gradation of some
devk.es 'that $tiU ha.ve \lsefd lives.
Due to -;onstraints in the models we used, there was "need to limit the number of DSM
pr~grams offered. v..'e worked around this constraint by packaging combinations ofDSM
programs into levels. More communication and feedback betw-een impler:'lertters ofDSM
programs and ptanners would ensure that an optional D';M resoutce set is used for
modeling. More levels would allow us to fine tune the size of the DSM. resource A
possible method is to sele<;t I. coarse grouping <:If levels and contin-uousty fme tune the
size of the levels in ~'1 iterative optimization routine. Alternately, we (ould also perform
t....,s~ ltltv7!t.d <JruJ idtruj",. imp"~\ii"Z t11~ }994 IRP proct.n
City of P4lo All" 19911RP "'-3
-.
stand-alone optimizaticn ofsuppty and DSM reSOUlces prior 1"0 resource-integrafion !:Ita,ge.
We should review l~e method of reducmg technical potent'icil to achil!v4ble potential
Measures used. 10 determine ach~evabll! potential should be consistent throughout the
process of redueing (rorn lec.Mical to achievable pOlentia! We should ex:amine the
factors ~d for reduction from technical to achievable poteraial fot further impro-"'ement
and verification. We should re\riew literature and evaluate computer models ~or
ttdmiques used to reduce tec:h.'lical to achievable potentia.! and also for data from similar
utilities to aid in this task. Finally. we could also formu!ate estimates .... ~t}. real life data.
In this IRP we examined only envlronmental externalities associated ""ith air pollutants.
Punher, we used CEC values and had no idea how Palo Atta!ls would value
environmental exte~a1ities. To improve this situat~on, we must examine how our
customers value ~nvironmental externalities We could s;:nd surveys to custome.s asking
them to quantify the percentage increase in rates they would accept to minimize
externality values. We cQurd imptement green pricing that allows customers to
voluntarily subsC!i.be toJ ntes that are, say 10% higher. in order to fund clean alternatives
above and beyond w~at we may have otherwise acquired. Wt. should also examine costs
associated with dtpriving a custoJmer of service. i.e. altering ]tfe style .. 0 the degree that
deprivation is fert
This IRP considered only air pollutant externalities We could expand the scope of
externalities to consider all aspects of environmental externalities. We should look at
positive externalities including: Jocal economic multiplier concept.. which measures the
boost in locaI·e;onomy due to building local generation or implementing a DSM prognm.
A.4 CommuDication
The IRP is developed for Palo Alto's planning !l~eds and is also provided to different
agencies ami depanments 'Wilo 'Will use the doc1Jm~nt for different purposes The IRP
should've developed in accordance with Western guidelines to preserve the integrity of
.our Western contract The IRP !hould be developed periodjcally 10 satisfy the CEC
regulatory mings. The IRP should serve as a guide to NCPA 10 clearly indicate Palo
Alto's Mon-and long-term plans Finally, the IRP should coincide with bt.:dget cycle
scheduIes to faeilitate management, UAC and Cc·uncit approval processes:. This implies
thai the ::l;ext IRP should be completed by December 1993, in time for the 1994-96 budget
cycle.
The IRP can be developed oniy with the cooperation of various div:isions. The scope of
M IRP requires data from l'1.umerous sources and the experti::e .and input of experienced
perscnneI from different divisions As a result, the:re is a need for interdi .. isional
cooperu.ion. Buy-in would be fostered when eai:h division is informed of their roTe and
responsibility in the IRP process and when the IRP time line and schedule is responsive
UUC1'U k.cJn1td anrf irf~(U for ,'»pro"ing rllt 1994 lP.P proau
Ciry oj PolQ A/ro 199; lRP A·4
,-, .
·'
,
)
to the time <;onst,air.u or t.a(h partic~~ating dl'\rision, InterdlvisiclOal cooperation cou;d
bew.anc~1;\ by sharing fe$Wts with all divisions. This eould be accomplisbed \1.Ilth a -d.""y
run" ~rl'ormancc of the monthly UAC presentation to a i:rOSS section of divisions,
Documentiog each step of the analysis process forces us to lNrite out Ol..!l assumptions a.nd
interpretation ()f the o'U.!puts, which witt d:iscoura,gc deo;::isions made on t..'1e basis of first
lm';H'tsslOnS. Keeping a. p~er tnil of the a.."\a1y~is steps used i;"l .each stage would also
belp us understand .and commWlic.ate the shortcomings of our analysis S)'stematic
&:»..unentation of r~su!ts will result i::! less time spent searching for data for UAC Cor
Council presentatioflS.
u1SdN' U4ntf:d.:lM drOJ/er improvil1g d!~ 1994 fRY P'QU.Jl
CiJ)' o/P(J1QAllo J~91IRP A·j
•
•
APPENDIX B DSl'>IANAGEn
The Electric: Power kesean:h Institute (EPR!) developed DSMar.zger :0 help analyze DSM
programs. We used DSManager to SCreen DSM measures and to gro'J'p me4Sures into progr2.."Ils.
The programs were tested for ecClr.norr.ic viability. The data describi11g each p'!'ogram was
expon.d to MIDAS.
Figure B.l sho'WS a. s~hemaric representation of the anaJytic frame ..... ork in DSManager. The
OVERVIEW OF DSMANAGER A:-;ALYTIC FRAHEWORK
FigureB. J
status of the utility and the customer v.ithout the program is compared to their s:arus with the
DSMtllUJg~r
CUy 0/ Pal" Alta J 9'}] IRP B-1
•
program.
Customers bave several categories of costs They buy en~rgy, both electric and non-electric:, from
utihries. 'Thet~"&n re:c~ive reba.tes on 'thetr electri; bill by ?artici~ating in the DSM program.
They may also incur capital and operaring co~1s for new equipment DSManager models the
imp3cts for parricipating cu!:tomers &nd the utility
DSM&nager uses the concept of "end-use-to organize customer energy use ."'u't e:nd~t.lse is a
speeific en~rgy ~rw1~e desired by & customer (i,e. a warm fDOm). and the cuS10mer cal'. seiect
different technological options (i e, e1e,"rric space heater, gas furnace. ct·;:) to pro\.i.de that se["";ce
DS~{anager uses these end-uses to define the loads !lnd e~ergy consumption befor.e and a..fter
DSM programs.
DSl\o!.anager divides the yell into seasons .and representari'/e days. Day repre£entatior.~ related
to weather are then grouped as day categories. and are used to capture variations in tempenture
stnsirive loads, mainly spa::;e conditioning. Day representa.tions related to human activity are
grQuped by day class, and are used :0 capture events such .as wc:ekend businesses closures
&pecifyiDg I. season, day ute-gory and day class defines a representative day type Energy LI...~
is specified by day type in D SManager.
The prices charged for electrictt)" are.a key factor in determinlJtg the e-o:onomic benefits a..,d costs
associated y,,-iU! a DS2' .. I program. A! customers cha.·,ge their electricity use and patterns, the rates
translale these chao.1ge.s into doUan. that impact customer bins end utility revennes. The rues
module ean model dlfferent rate ~chedules .t.CTOS$ the year.
DS.Manager models :he utility costs· both fixed and variable -or the DSM program. We can
specify either one-rime cr annual costs, and both can either vary 'W'ith cr be independent of the
number of participating customers A."lother cost category is yearly utility sa.,ings from peak
eltctricity demand reduc~ion. A unit reduction in peak demand reduces utility costs for
generation, transmission an~ dlst."'ibution equipment at a user-specified ratc.
The utility also saves the .... ariable cost of production wb~ electric use declines at any time
during the ~'ear. These savings are mainly fuel cost sa.,~ngs, and they var), significantly across
the year. These cost s.ayings can be caI..:ubte.d using detailed production simulation mod~.rs, but
DSManager uses a. simpler me1;od for representi"g these savings by specifying a marginal cost
for ea,h hour of the repres!ntative day typ~. DSManager uses these marginal costs and the
detailed chUiges in customer loads to calculate the change in the utility's ..... aria.ble cost of
plodudng tiectricity.
DSJ.Jc".£J1I~r
Cily of PaY; Al.o } 991/RP B-2
•
.. ..
... ..
...
...
...
'" ..
•
• ..
•
•
•
•
•
•
•
•
•
APPENDIX C MULTIPLE-OBJECTIVE
INTEGRATED DECISJON ANALYSIS
SYSTEM (MIDAS)
Epri's Multiobjective Integrated Decisio~ Analysis System. t-.1ID..4.S, i,~ a planning tool that tackles
the Lmcertainty a.spect of the declsion making process by prov\ding .. decision tree framework to
quantify and conceptuahu planning issues The dt'Cision tree frameworl injects risk into the
&rI:l1ysis from the onset ~1IDAS is comprised: of four basic moduJes --decision tre-e, simu!ation
system. resrJJt:S, atld dedsion analysis:. Each module is di~ussed separatel~· A detailed
description of each of tiese modules is available in the MlDAS documentation
C. ! DetirioD TreH
C.2
The decision tree is the framework for organizing the many s,enanos to be analyzed
Each scenario or endpoint on the tree representS a potential state of the world 'With
d~isions and 'olJH::ertainries .
The decision tr-ees CaJ'l have approximately 600 end-points and can take any shape. They
do not have to be symmetric, and phased deciSiotls c:an be analyzed by alternating
decision forks iIlld chance forks. For each branch in a tree, any set o{ in.put variables can
be changed. FtiI cha.,ce br4llches, the probability of the event 15 !pecitieo. The
information on the tree is used by the slmulatioD module 10 produce results for each
branch in the tree
This syst~m integr.tes load analysts, capactty planning, proc'uction c.ostir.g, finance, and
rates to simulate the operations of the utility in detail. MIDAS's simulation system
functions are ron for each year of the study before going to the nex.t yp.ar of the study.
C') MIDAS Load Anal,:.sis
C.4
MLD.<S
1YlIDAS projects monthly pe.1k day. weekday and weekend load panems based on
histcrica! load patterns and peak and t'!n.;rgy for~casts. DSM programs 'With different
objectives (i.e. peak shaving. load shifting or conservation) adjust these p.ttems. Load
shapes and DSM impacts can be e\'aluated at the sys;em andior class le .... et
MIDAS Capacity Planninz
The most straightforward method for adding capacit)' is to specify the completion date,
capit1l cost. capacity and operating costs atld cbaracteristics of the ~nit Using: dedsion
rules.l\fiDAS v.i.!! !flake capadty additions from a set of options. The analyst describe~
generating units to three capacity mix: or operating categories --base:Joad, cycling. and
City rJ/ PQUJ AJI(j /99] lRP c-!
"
--.. . -
peaking Planning ratios ref1e~tin& the target mix of these W1it types are specified.
MID .... S .....,jll add capacity when targe: ratios are not met. The user can specify the degree
to which peak load red'Jctioll by DSM prcgranls is co·.mted as a capacit:l source. These
reductic.ns in peal: demand can displace the need for either peak, cycli:lg, or basefcad
capacity. This helps to evaluate the trade-off benveen :illpply and demand program
options
CS MIDAS Produdion Costin:
Generatins units in MIDAS ate classified as energy·limited, capaciry-limited or purchase
contrac.ts. Energy-limited units .are generally hydroelectric run-of-river, pond SIc-rage, or
pumpe<J: storage. However, any tra."1saction 'W'hich js r;ot dispatch.1ble, dispatched for peak
sh .. ving., dispa~ched on-peak or paid back off-peak can be modeled ~ Ul enerIT-limired
unit The energy expect~ from these units is used 10 modify the load curve. Energy
from run-of-river units reduces load in ail hours of the month Energy from controlled
Wlits is used to reduce peak. The energy input needed for storage! operaticn fills the off
pw bours.. The c.apacity·limited unltS and purchase contracts are dispatched Igainst this
modified load curve. The user can choose between three algorithms --probabilisU;;,
deterministic, or a c-o:nbination of the t"wo·· for the treatment of forced outages when
calculating productio[J. costs for capacity·limited uni!s The trade-off is wtwecn
oomputatiot'l time and theoretical accuracy.
Purchase contracts are trealed deterministically and have no forc~d outage rate. Contracts
can be modeled with a variety of constraints, including minimum and max..imum energy
ar:d capacity takes. W~ moceled the majority of our resources as purchase contracts
C.o$ MIDAS Fjnanc~ and Raft's
MIDAS
The finance moduJe in ~AS was customized to reflect Palo Alto's finandal model, and
binges on Palo Alto'! res-erve system, ?t.finimum. maximum and :arget levels of reserves
can be spe;:ifled. The model will caJculate the revenues coEected each year depending
on tl,e rate, In turn, the j.&.te can be llser specified for any or all years The mode! can
also cakulate the rite each year based o!:. cLr.ain constraints on ... ;hieving reserve level::;
Rates will be calculated to generate revenues :0 meet minimum, target or maximum
reHlve levels. Additionally, th~ rate inl,;rease per year can also be constr:lined. Finand:!..l
statements can be generaled from the results of the analysis.
MIDAS calculates customer-class average rates usin.e; a simplified cost allocation. The
user can provide input I" ctassify base revenues to demand, energy Uld customer
categories. ~nDAS uses customer class energy, peak demar.d a...,d number of customers
to a1loca.te these amounts to the customer dasses and calculated average rates. If price
elasticity information is available .and the user chooses:, l'¥fiDAS can use thi~ price
infonnaticll'l and user specified short-and long-.erm price efasticities ta ml}dify energy and
City o/PawAIIo 199; lRP C-2
,.
; '.~'
"
"
•
C.7
C.B
MIDAS
..
peale fore"Ca.s~ for the-ne:d year of the simulation.
MIDAS R .. uil!
The massive a..'t1ount of information prodi.i~d by the simulation module is organized into
a "'results box.· The user can !'ipecify any subset of the 240 variables, years (30
maximum) and scenmos. Th esc values are displayed on the screen either by scenario for
III variables .and all years; by variable fOf all yeats and aJ.! scenarios; or by year for all
variables and all seer.arias. The results can be pri:lted or sent to a file (or use by a
scp.arate spreadsheet or graphing program to allow additional analysis andlor graphing .
.t-.flOAS has detailed reports that the user ca.., turn on and off by year and elldpc.int
'These reports pro"';de infoImation on mOl',rhly or annua.! pr..,dl..!ction by gener;iting unit,
monthly load shapes, expansion choices, and re\~nue aLlocation for rates.
MIDAS Decision Analysis
The decision ana.l)'sis module in: :MIDAS is used 10 develop obj~tiye functions and risk
profiles. In spe,ifying the funC1lon, the UseI C&l apply any of several operations
(minimum, maxlmum, average, sum~ or pr~ent value) to any OUl'put varii:.Jle-or
combination cfvariables. Multiple objectives un be combir,ed into compound obje"Ctive
functions. Some examples of ubjective functions are: present value of revenue
requirements, cumulative SO .. tonnage, and levelized system average rate. Compou."ld
objective-functions help the user develop i!liighrs into risk eJq)osure and rislc management,
thereby allowing the development of strategies that are forgiving ir'! the face of
uncert.a.inty_
Ciry '" P"lc Allo .'99) lRP C-3
.'
APPENDIX D DETAILED TRC & SAR LISTING
~I ~I~
I :1~h. L
1=
t J ,: ~
I ~ ;:-
I"'" J"
I"" :r.
III I I. III
II \C I'
.t: l;i 1" 0
U ::;, ..
I.. 141
1.3 I~I :n
It I" ~ <, !;,j ,}II
I: H.l :tI
~ I:!': Jl
• " I~ :I:l I~. ;Jl'
1:: ,. I'
:a ~ Il ,
" ,
•
• • • .. • • ..
0
:1' ,: :;.. I>.;
He Ic;I
... : l:
~I:
~ " :~ j :-
:. II: :. ,
'" )01 D
:. IJI !.!3
.. 1100 ::10
.. 1:. ,Dol
.J! I=-1J
G lJ~ l<
.. • I:' · '" • JaI .".
• :I: II o
I~' II: .
:II :I" .-:
D Sl )l~
SO III:: .):;I
ti ~ "
• "': r 0
,., ~ ':r.
• ,.,. .Jo
lit !.OO :l!
.. 1<+ ::n
.. :t). '"
til Ul' JI-
til 1<': ;:n:o
II D) <::: · -~ ... llol ~
V.l4iud tRC &: SAR lisril'lg
City of PQUJ Alte 1992 /RP
• ,
0
0
• • ,
• ,
0
•
O
• • 0
•
0
0 •
0
0
I ,
0
0
0 , ,
• , , , ,
0 • • • ,
0
• , ,
• •
0 •
0 ,
• •
0 • , , ,
• , ,
I • ,
•
~ , ,
" .. , ,
0 ,
I
•
" " , ,
I " I ,
" , , ,
• b • • , ,
" I • • , ,
• I • , , ,
• • • • , ,
• , , , ..
• • • • •
" " •
" • • D
" •
" a
" a
a
" , I
~'f~-.:n~ I,~I F"(10.~
~ ~ ~t.'" ~ 1" = • 0 0 0 ~.
• 0 0 , ; ~ "".110 • • 0 0 .or · • 0 0 · " ..... "
0 0 , , · I " .. ,,.
0 0 0 0 0 u ",. · • 0 , , I " -,
• 0 0 0 , " ~.
• • 0 0 , ~ .," • , , • • " .. III
• • , • • u u.n
• 0 • • • u ...
• • • 0 , u _.
• • 0 0 • u ~.
0 • • , • U .or , • • , • u .o~
• • • 0 • " .. 0 • 0 • , • U ~.
• · • • • u ~ .. • 0 • • • u m~
• • • • • u m. • • • • • U ..... 111 • 0 • • • C ~" • · • • • U Qt,'"
• • , 0 • U ",I)':'
• • , , • u .. ~ :::
• · 0 , 0 U "'II: • · • • • u .. * • • 0 • • U H,'
0 · , 0 0 U "II.
0 0 , • • u . ..
• • , • • " "".Ii' · • , • , u .0' • • • • • u "': .. • • • • • ,. ~" • • , • • U m' • • • , , u ,,~ 0 • 0 , • 0 , u .0" • • 0 0 , u ... ,,"
0 • • • • , mo,
• • I • , • " -': • , • , • U IJI" • • · • , u ~.
• , · • , u m.
0 0 0 0 , U <J'''' · • · 0 0 u 'no
0 0 • • 0 u ~
• • • • • U "'\1 • 0 0 0 • w ~.
0 · 0 , , u ~r.
0 0 , , , w .,.
0 0 0 , 0 " ~. · 0 , 0 0 u ~ ..
• 0 0 , , U --,
0 • , 0 , " .,.
• • , 0 , U "'L: • • , • • u -. • • • · 0 " -. • , • • • U -~ • • , 0 • U .'-• • • • , " .'"
0 • • , • " _.
• 0 • , • 0 -. • 0 · 0 • " --• • · I 0 • " ~"
0 0 · , • " .,"
-1 ~~ r::-
~L;'
~>
T;..I' -,.
"" I "-"
I 7",.,
0I.r.
"-"
"" I '" ..
""...ll
"" 1l.r.
"-"
'0
D-
.~ ..
r..,-,
r~ ..
1<::
r_,~
,,~
"T~'"
r._u
••
,,~
:"::' ..
7:.
'r;1J
"-" ". n'
"" "~
1:..I'l .. " ".
".I:
I
r..;:
'.'f ...
l':'.J:
ll~
"" 'l::16
n-... ••
~r ,
r~
>:11 •• •• .. "
~" nu
"-"
~. ,.
nu, I n.
1;..,-
~
D-'
.~.
·.'
O~ ~ .
~~
~~
~~ ~ ~ -------------------------------------------------------------------------------~E 5~iiji5~Ui£ijEij~2Si~~~nE~iili~~i§~~~.~I~t~~~ ••• "~~nn~8~~d~dt~~~~'.~ ~
~ .~ ------------------------------------.--------------. ~~ ;~~~~~~.ifH;~.t~'i.8i~9~.~§tiE~U~~iSE~~~i~ln~~!~~U5Hr~~~E~~~ifi~~iea ~i
~ ~~i~~~~ii~~iSi=U~tU~M'E~8U~~~~llinRjn~~~liU!;~~~~~~~~~~t~3~i!gg.~~~ E§
....... '" '" _ ..... ,~ ........................... '" .................................................... ~ ................ ' ................ \0 ~~
.................................................................. .................................................................... "I~ E
................... ,.. .. " ........... ~ ............ " ......... w ................................................................ ~ ................. I~I ~ .' ~ il .............................. -....... -....... -........................................ -....... --la
C~~t~C~c~~~CC~ctC~C~~cc~c~~ccttttc~~Ct~C~CC~tcc~cc~c~~c~c~cccc~c~~, 1~
.. !' "-to t • l:: t ~ l-~ I:; ,., ."l t . ., .. t t; .'l; .. t ! to t; ,'I; ~ tit-t .. r t t r .... Itt.: II .I: ~ It· t t-.. to t .. I " I ........ t< t .. t t-\ ' ';:b~-·-::':,..I~~O.::~:::I ... -":I;;t' ..... ,-Io':", .. ;t~ Io':-•• ~f·,.t:: t"r;;-:,: ... :::'.t~Ii_,~ ~~ .I:~tt"l .. ~_,,.:1 ~
." oj t :! II ., ~ U ;! ~ \01 :r ~ II " fJ • tl .. Ij; • Lt :.0 11 '-' ~ ':l ~ '1 t! It :c W 'r: ~:: I_I tl :s ... 'II ~ R ";.' b .. '1 '1 ~ , t.: ~ r II It .1 IJ ,;; t It i' " II ~ ;0 n I"i Ji
;I ~ ~ ;:l' ~ ~ OJ ~ ~ jj ;1 ~ {J ;'! ;: 11 ;: Of il ;' <l >f ~ ~ tt ! ii ~ :: ~ ,J r. :t ";1 :! ;i ~ i' ~ i1 i1 Of ~ ". ~ >I iJ fJ if 1'1 r1 if :-1 ,I ~ ,1 ;J ;1 il ;1 ;J d " ;t <iI-H
!.: " , ~ ~ • l1 !l \! • ~ " ~ ... :j \I; :.1 ~ \I; ,j OIl .. 'lII ~ ij :1 :II .. II I " II SOil t 1:1 " .: ii ii !J J X ::. :I II 1 e; or ,J .; ,J U ;j If 11 ".: R .! ."1 !:. !"; t., t; ;"I to· j t
j
" "
, 'f\
,
'oJ ",--Y--
"
1;:i:
n .~~ ~ ....
a-)J :.." fiRoo
~~
:::~
~~.
?
~
.... '0".:.1 \V -..:....-..:.I ~ ~.
------------------,-------------------------------------------------------, i$i~i~S~iii~ifti~S=i~Ji~~?a~~~ii~&ai~a;liiSi~~e5~.ii~i~E;~~iisS¥tt~S ~
•• ii.i!~fU~i!~&~~~~I'·i!~FHII!6Ini~!i!n§.VI~.~!~i'li!~.!liij!~~Ii5~ ~~ •
~i~lE&tfi~!n~I~~~~~~ft~~;~&!f~~~~~li~.Hi.~~~j~~,~~~t.~~~~~ii~~~!i~iij Ef
A __ ••• M __ Y~~~_R_~~~~~u~~a_~_u~~~~ ___ ~~ __ y .""""-."-.--""---"."~~
i'i r. ;:) r; ;; ;:; ................................ -...................... -......................... -................................... "I~
............. _B ...... __ ...... __ •••• __ •••••• _ •••• __ ....... _ ....... _ •• u ................ f j
.......... _ ........................... _ .................................................. ___ •• a ___ .~ ~
U~~~_W •. a._ ...... _B~_~~ __ M_~~ ___ •• D~_.~ •... ,.U_W~_ .. U_~~~ __ ...... _ •• a.1 ~
. 1
_ .. w .... _ .......... _ ........ _ .. ~ .................... _ ...................... ~D B ............................. lO .. -......... i
--------------~~
_ ... www_.vW~W_ •• NW .. _.VNM __ ... NWW __ .WNN._.""~ __ .MM_ .. _.WWW_ •• NU __ ~.W~M~
CC~C4CCC~t~CCtct~~tr.~tr.~~c~ctccc~r.~r&r.~cr.ctr.~cr.cr.c~cr.cr.cr.tr.ccccr.r.~1 ~I
S SE H ! P " ! " I P ! , I \ , ,1 ! " •• '.H. ". Ii H U I , ~ ~ • II t ~ j\ I ,\ I 0 • " ! • In'., n H H ; HI'" ~ :t tl ~ It '" h ;,j • ;.! au" ~1 ii ~ 'I' " U '" }1 ;;; :;. it !: ;0: S .,.,,, II • :< n :;. \: :; :, If l' ~ k: :, ~ 1; !:I \r :51 a: ,J ~ I 8' ;1 '-l l! 'I :-t ;. .. ~ w ~ II il ~ Ii H
f1:! a tl ~;I;1;:j iJ j1 ~ ,f i1 ~!i;J ~ t1 ~.;t;1;I;1;1 rt:1;1 11 OJ tI r;J ~;1 ~:!;I tf ~;1;1:! j1 OJ:! ,1,"I::t j1 if.:!;J:1 ~ 11:! ~ i!:! ~ if ~:! i! ~ ;jE' !f
1,1 t .. I; 11 1: III A ..: ~ a '), it 0 ;J iii " ); II ., , .1 n ~ 'I .~ ,J • it !t :, it ~ .,:: ~ ~ " :1 " :.:: ~ ~ « i: II: .:l :~ •• 4:: ;J 'i: 'r Ii • If l' :, .. 1'1 ..! ): ~ :, _ J~~,_~
. "
. ".~~~/
, :i'
~ ... "" ...... "" ...
,. ~ " * • ,.
3' " "' ~ • ,
~ u •
~ U l~
'"' ~ " '" • ,"
~ • '" * • • .. u •
'" : '" '" • 0
'" " • ,,' " '" C' 0 ,
"' 0 " '" , ,u
'" I
.. • ". • "' ". .. ..
~ m " ., • JI:
'" • e,
'" ow • •• 0 •• .. 0 ~
'eO ~ ~
'" " •
". • .,
'" c, 0
~ " '" ., 0 '" '" m " '" • •
'" • ,. ., '" »
'" • ~ .' • '" ... '" • '" 0 • ,.. • "' W ... 3 ,. m .,
'" • I '" '" m •
'" I ~ ~
~ • ,.
'" ~ " ~ -'" ~ » ~
~ '" " ~, '" ,n
"' • .,
m ,. m
~ ... ,. ., c ~
~ '0 ~
m '" :n ... ~ -m ,. , -'" ,u ., • ,~
~ .. " ~ ,. '" '" , » .k
D~tailc.d TRC &-SA.R liJIi",8'
CIty af p(J/() ..tIro )99) IRE'
"" '" • " • " • c
• ,
0 • • ,
• n
I • c
I • c
0 ,
0 u
• " • " • " 0 c
• " • " • " • c
• c
0 c
• " , " 0 " 0 c
• c
• c
• c
0 " 0 c , " I • c
• c
• c
0 " 0 c
• " 0 " • " • • • " • c
• C
0 ,
0 C-, • • c
0 c
• r.
0 " 0 " 0 I c
• n
0 " • " • " 0 " • " • " • " 0 " • • , " 0 " • c
• "
.J;t.'H'\..Y.U'JL'Ill:Dq-,
~ ~ """" .IJo1SIY ~=
• • , • ,
• • , 0 ,
• • , • ,
• 0 , 0 0
• • , • • • • , • • • • • , ,
0 I • 0 , I ,
0 0 • , ,
0 0 • , 0
• 0 • , • • • • , • • 0 0 • ,
0 0 0 • ,
0 • • • ,
• • • 0 • • 0 • • ,
0 0 • • •
0 • • , ,
0 • , • ,
0 • , • ,
0 • , , ,
• 0 • , 0
0 • , , • • , , • • • I • , • •
0 • , • ,
• • , 0 0
0 0 , • 0 • 0 , • •
0 • • , • • • 0 , • • • • , ,
0 0 • , •
0 • • , 0
0 0 • , • • 0 • • ,
• • 0 0 ,
• • • 0 ,
• • • • • • • 0 0
• 0 • 0
0 , , ,
• , , ,
o , , ,
0 • , 0
• , , • • , • •
0 , • ,
0 , • ,
• , • ,
0 • 0 • • , 0 0
• , • 0
0 • , •
0 0 , ,
• • , ,
0 0 , 0
0 • , 0
• • I • , •
0 0 0 0 ,
0 0 • • ,
• 0 , • ,
0 • I 0 0 • • 0 0 0 0
• • • , 0
-~"' . ~
"'" " ~M
" I "'~
~ I .:,;: rl
" "' .• , u .,.
U 0)1"
w ~*
" ~"
U 'l'~
u "''' u «.
I
U 'l'~ u .....
u .~,
U 'Jl?'!
U QU"
u ...
" "l,~
" ... "u
" "lUIO
u .It.o
" -" " 'l1..loi
I U W£
I w -..
u a''''
U 'll",
U ~
u .0.
U 411 ""
U -W ~}1 !l
u "' ... " -~ u ~.
U .w
" -~
U .~
u m.
w -,
u ...
U "". U ~.
u ~". " <J~:.:o
" I "'~
U "'loll
U lla :1
" .u;.
co ~~
u ... ~
" ... t'l
u -~ u .....
" "'H:
u ' ... ~I
u .~~
u .,.
u •• n
U (;l',;]
" ~~
U .... 1 ..
u ~"
" .,~
u .,.
U .Q"l
-l ......
~
Me ". ~ -"" 'l':.ll
ftC
~I,.,
"~
~"
,~
r..1:
u
.~
:0: I!
U
"" ,,~
lUll
"~
:0:." ".
11'"
C,
1;.::
I ".
"~ " ... .~
C" I ",.
~; ..
,,~
co ...
1'!1:
n,jI
U
1': ~l
,:.". ..
'O-
.~
"" r-r.
."
! =,
r.::tl
?I/.'1
u
1'1:':1
110 ~ ..
~u,
"" 73.'6"'r
"-"
'Ln
"" 1:11 .,.
~~
'1':.1'
"" ."
I co
1"),1"
D-4
,
J
.
"
•
.....,.. ~~
'" ~I
'" '" I .. ,.
'" !:I
'" '" ,. ...
'" '" ro ~
'" n'
'" " n. llt
rn ••
~I •• rn 'l~
~ '" m no
~ ~
"' J » ., '" m "' • rn
'" • -* ~ ~ ,.
I ~ ,. .. .. ~ .. ,.
'" '" m r
~ " ~ c.
~ ~
~ " '" ~
o. • -" ., W I ~ •
"" • * •
* • -a -• -,
~ " '" • ", ,
I III ,
" or.
0"
,.
'" ~
'" 00
II' ~
'" ~
'" -m • n, r-
'" '" = • •• 0
'" m .. ~
'" • ,.. '" '" " ~ •
D.liJilrtl TRC J: SAR listing
Cir)l afPala Alto 19921RP
m,
"'I' '" , ", .
" I ' :1 :
U,I I G .. ,
'" ,
~ • m • C, • 'W I •
31. • n, •
C-O
"0 •
"' ,
'w •
~ • or. •
'" I •
", • on •
" •
" · .. • • • .. I • 1\\ • n •
" ,
u:. • • • • ,
'" • n •
" , ,,, • , • • , ,,, • , • • , ., 0 , ,
" •
,~ •
~ • ", 0 , .. • ,. •
I::: , -•
" •
'" • ,. ,
,~ I ,
• ,
" •
" •
'" •
,~ •
ll~ •
,~ , ,. ,
~.
" " " " " " " " " " " " " " " " " " " c
I " " " " , , ,
• •
I • , ,
• • , ,
• • • • , , , ,
• • • • ,
• • , ,
• • • • • ,
• ,
•
I • • I • •
•
. .
R.~'tll!*ll."-CZ:S~ -""('1.:1] ~ « ..... ..."",...
~ 1Jr.....,~ ~N ~~ W '"" , 'IT 0 0 " -" 1 ..... , , , , , ~ ..... !': 1"'..!. , 'I ' , 0 ~ .....,0: ~ , , 0 , • W -" 1':. . .1' , , 0 , , W .,1'" ~, , · , , , U .d :~ I :-: ..
• I ' , , 0 " W6 :~ 7;F , ,
I
, , , I ~ ...... c: , nr. I , , , • , ~ ':!II." 1"l...,-
• , , , , " _D ", , , 0 , , ~ .~r :0: t~ I • , 0 • , '" ~" "-" • , • , 0 U ~" :'Oal I , • • 0 0 W ""J r..r. , • , , 0 U "'~ '1""....JG
• , • , • I " ~. '!"L:'!' I
• • • 0 • W ,),0:
I
~"'I , • • • , W .~~ r.·1 , • • , , ., .. \\,. r ••
• , • I • • w ~~ ~H , , • • • u .u.;:. ~~ , , • • • u u.u: "~ , , , , , U ".JllI ,,~ , , , • • u -" = • • , 0 , u .,.., ~" • , , 0 , w ':If:1j" 'j~ , • , 0 0 u ~~ n" , • , , , W ~,II '1"'~ • • , , • " ••• 'D , , 0 0 • W ."11 "-~
• • 0 • 0 " -. ~. , • 0 , , U ..." no
• • 0 • , u c~ ~., r~
• • , , , u .,~ ~ , , , , , u ... "~ , , , • • U Q)ll 'lmo
• • • , , " -" -• , , , I • u -.. u
• , • , I 0 u m> ,,~ , , , , , u . ~ •• , • , , , w -~ u , • , 0 • u m" ".
• • , , 0 u ~" •• • • , , u ... u
• 0 , 0 U 'l']1 ...
• • , , u .00',;111 I "" • • , • u wo ••
0 • • w ~UCid 11.f:. , , • " ~. "* , , 0 u ~. ,,~ , , , u <on "" , 0 • " ,j:'~ ". , , , U .... 10 n:n , , 0 U (.1'''''' "" , , , " .,. nr. , • 0 " 11':' . .101 11.,.,
• , • 0 W -1.0','" ,,~
• 0 , , " d1.n ". , , , , u _c u
• , • • U 011 a< ". • I , 0 , " .~,;:, ".4
• • , 0 u ol')" "'~I
• I • , , U <"11 "" • • • , • W IIl.J'l "11-1' , , 0 , , " ...,1,1_
I " . , , • 0 , u .. u. ,,~
D-S
,c-1
i~l;:
"b ~-t
~;.:
." ...
~~ ;on
~a.
~~ :::'" :::~ ~~.
?
'"
.&!~.J~!~!ISa!l.i!I~~~5!i~i5~1~~E!~5~![I~;II[i~!~I~.EI ~
~l~~Y~~~n~ij~~~li~!!~~i~~~i~~~R~~~~U~~~~.~~~~~t;g~r~~~~ ~§
~iiH;g~»~W~g~~~~;~!~~~~~i!a~!"~~~u~a~~~~~~~l~~~~5~E~1~. E~
I" ........ D .. EO .... D ........ n ........ De", ................................. It .. <> ...................... "l~
.................................................................. ".. ............................................... I~
~ 1 •••••• ______ •••••• __ •• __ ••••••••••••••••••• 00 • __ ••••• ~~
_ ... ~~_ ••• _NN ..... _~N ...... ~~~ .. _ .. NN~_ ... N __ ..... __ N ... _N"N. __ ".~I~
EC l;!: C r. r. Gr.!: t: r. ': C r. ~ c r,1" t r: t t r,!: C r. t (; r. I: r; r; r:: c r:!:' C" r.!: I: r: c cr.!: r. t t' r. (,:' tl ~f
; ~ i II! ! t 5 l ; S ! ! ~ ~ ! ~ f t ~ t r. e t e s ~ ~ t S ~ ; ; ~ ! ! ~ £ t i ~ ~ ~ ~ ~ 'i .~ ! ~ 2 5 .g 2 5. ·I~ ~ ~~~~~~~U~JJ=k~~8.n~~G~~~~J!~b~"a'~at~~e~"~R~S~_~~~~c~~ J r ~ ;1 ~ t\ ;:I ;;2 oj ;:I fJ i! .:1 t1 ;j ,1 ,I j;J ,I tt ;j 1"1 r::t ;t if ;t pJ :! ,J ,I 01 ;1 ~ ~ 11 ;:J <II'. ". ~;I Of it ti. H ~ i! t1 iI ~ t" ~ i~ :I :!,<
!!I ::.; ~ !': .: ~ t: ;; ':1 ~ III !;: I!l i ~ tI ;:" ~t ~ ~ ~ .; ;;I w. ., -!:. It ~ ~ n. w ',A • ~ !::I iI I 1\ a I.: ~ ~ ~ i; :::: r ~ 't \Ir 'j ~ II II :. ,. H
.
:'
,',;
, ,
.~,-
'f:,
~