UTIJTY SYSTa1S ENGNEERNG DEPARTMENT · lr J . • 8 ECTRC UTIJTY SYSTa1S ENGNEERNG DEPARTMENT Document Number ' d5.37 I Please Return To /J4 ~..:7 5" . DOCUMENTCONTROL /7~/ PROGRAM

lr . • 8 ECTRC UTIJTY SYSTa1S ENGNEERNG DEPARTMENT

J ~~~~~@ _________________ _

Document Number

' d5.37 Please Return To /J4 ~..:7 5" . I DOCUMENTCONTROL /7~/

PROGRAM DESCRIPTION MANUAL J f

I (1

It !'

Jf

Jl

~~

~i

·Jl

~i.

Jj

~

L

MUL~AREAPRODUCTION

SIMULATION PROGRAM

PROPRiETARY INFORMATION

GENERAL ELECTRIC COMPANY

VERSION! 3 JANUARY, 1984

GENERAL. ELECTRI:

I -· _,t

I !

~ i

r,. . . '

I i

i ;

I I I Chapter

I 1

I 2

-- 3

-I~ - 4

-~

-I

-I I I I

·-~· I << I

\l

MAPS PROGRAM DESCRIPTION MANUAL

TABLE OF CONTENTS

INTRODUCTION . • • • . 1.1 MAPS Scope •.• 1.2 Who Needs MAPS?

. . . . . . .

. . . . . . PROGRAM DESCRIPtiON . . . . . . . . 2.1 Overview . . . . . 0 . . . 2.2 Key Capabilities . . . . . . . 2.3 Input Data and Program Size . 2.4 How MAPS Works . . . . . . . . TECHNICAL DESCRIPTION OF PAR'l' 1 3.1 Load Model . 0 . . • . . . . . S.2 Loa~ Kod1f1~ations • . • • 3.3 Exter'nal Contracts . . . . 3.4 Inter-Company Contracts . . . 3.5 Part l Dimem· 1o11s 0 . . . . . tECHNICAL DESCRIPTION OF PART 2 4.1 Main·r.enance Schedul' lg . . . . 4.2 Thenr~IS.l Unit Data . . . . 4.3 Transunission Data . . . . . . 4.4 CoDIIlitment . . . 0 . . ~ . . . 4.5 Energy Storage . . . . . . . . 4.6 Dispatch . . . . . . . . • . . 4.7 External Contracts . • . . • . 4.8 Cost Reconstruction . . . . . 4.9 Output . . . . . . . . . . . .

i

,. b '.

. . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . • • . . 0 .

. . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . 0 . . . . • . . . . . .. . . .

• . • 1-1 . . . . . . . 1-l • • • . • • . 1-1

. 0 . 2-1 . . . . . 2-1

. . . . . 2-1 . . . . . 2-2 . . . 2-3

. . 3-1 . . . . . . 0 3-1

• . . 3-1 3-2

. . . . . 3-2

. . . . • 3-2

. . . . . . . 4-1 . . . . . 4-1 . " . 4-2

. . . . 4-3

. . . . 4-3 . . . . . 4-5 . . . 4-10

• . 4-10 . . . . . . . 4-11 . . . . . . . 4-12

:I .,.$

I

I PREFACE

The information contained herein is proprietary to General Electric Company and shall not be used, duplicated or disclosed in whole or in part except as authorized by a written agreement with General Electric Company.

Copyright ~ General Electric Company - 1984

ii

I .,

I I 'I I 'I I J

--

.. ----.

-----

CHAPTER 1

INTRODUCTION

1.1 MAPS Scope

MAPS is a powerful simulation tool that accurately models the operation of a multi-area electric utility system and determines fuel use and total production cost. Represented in the model are the unique characteristics of each generating unit, load diversity among different areas of the system, and limits on the transfer of power between a~eas. MAPS considers transmission limitations that may restrict the free flow economic dispatch of the system. The commitment and dispatch of the system observe all the transmission constraints . In a pooled configuration, some utilities will be buyers of energy and others will be sellers in any given hour. MAPS can calculate the savings made possible by integrated operation and the allocation of the savings to the member companies. 'I'he MAPS program performs a detai1ed hour-by-hour simulation or the utility system's operation.

1.2 Who Needs MAPS?

System planners will find MAPS useful for generation and ~ransmission system planning, financial planning, fuel budget preparation, maintenance scheduling, minimum down time studies, and marginal and av·erage system cost studies . MAPS can be particularly useful for any production costing study that requires:

o Representation of limits on the flow of power between areas.

o Representation of optional external contracts.

o Determination of savings due to joint company operation.

o Allocation of operating costs to member companies.

1-1

I I I I

CHAPTER 2

PROGRAM DESCRIPTION

2.1 Overview

The MAPS program has two major steps. First, the program determines the minimum cost dispatch for a system containing transmission limitations, either between areas of a single company, or involving the service territories of many companies. These limitations are physical restrictions to the free flow of energy, and may not necessarily coincide with company b!>undaries. When operating in this system dispatch mode, the MAPS program determines the impact generation unit or transmission line additions, changes in ioad growth, costs of fuel, and changes in other system parameters have on system production costs.

Second, the program determines how each company would.have operated if isolated from the other companies. B¥ comparing the system dispatch to the isolated dispatches, the buyers and sellers of energy can be established for each hour. Company costs are calculated which properly account for the savings due to the integrated operation of the separate companies. This calculation is kno·wn as cost reconstruction.

2.2 Key Capabilities

MAPS has the following capabilities:

o A chronological load model allows seven differ-ent daily load curves for each two week inter-vul for each company or ar.ea. To allow for diversity, each <:ompany or area can have its weekly peak occur on any day of the week.

0

0

0

0

Thermal units are dlispatched on an equal incr-emental cost basis using multiple stepped incremental~ to repres1~nt the thermal characteristics of each unit.

Thermal dispatch can ·ae: Deterministic. Deterministic with deration by forced outage rate. Probabilistic.

Hydro can be represented as~ Run of River. Pondage. Pumped Storage.

Contractural purchases and sales can be modeled.

2-1

I

I 'I' ,I ·It . :.1

I

I

,.

0

0

0

0

0

0

0

0

0

0

0

0

Automatic maintenance for thermal units can be scheduled to recognize the cycling nature of unit maintenance.

Manual maintenance can be specified if desired.

Forcad outages may be modeled using random maintenance.

Spi.nning reserve requirements are me+-. on an area and company basis.

Data input may be provided annually or on an interval basis.

The program can skip the production cost calculation for intervals or yearly periods while updating the system data for any changes during the skipped periods.

Present worth of production costs can be calculated.

Forced outage rates for each power section can be considered for thermal units.

Wind and solar genera~tion can be modeled.

Output summaries art!! writ ten on an interval, monthly, and yearly basis. Outpl,tt includes summaries of production cost. generation, plant, and fuel. Production cost summaries a':"e available for the pool, areas, and individual companies. Production cost summaries are also available for the isolated dispatches of the companies.

Capability is included to calculate marginal cost and average system production cost.

Customized output options are allowed. for programs users to write their own programs to create special output summaries.

2.3 Input Data and Program Size

The MAPS program requires the following input data:

0

0

Load Model - The load model is a chronG~ogical sequence of the hour-by-hour expected loads for each area and/or company in the system. The load model can be calculated from historical load data using a General Electric load modeling program. Currently, MAPS can model up to 11 areas and nine companies.

Capacity Model - The cap&Gi ty model describes the performance and cost data for up to 225 generating units. Every unit is modeled with a minimum, continuous, and maximum power output, along with any number or intermediate power points. A m~n~imum

. ·•

~

(,

"

.•. ;~

'" .. . ,

;)

I I I I I I I I I

-I I

---~

~ JK .~

I

of 900 incremental loading sections, divided in any manner among the individual units , may be specified .for the system. For each unit, the fuel input at minimum and the incremental heat rate for each delta section must be supplied. Additional unit data includes fuel type, fuel cost, minimum down time, fixed and variable operation and maintenance costs, maintenance requirements and forced outage rate. The area location and percent company ownership for each unit must be specified.

o Transmission Model - The maximun1 power transfer limits between ~eas is specified for both directions for each hour of the day.

o Load Modifier Data - This data includes the specification of the operating rules for all hydro units and contractural sales and purchases . MAPS can model up. to 20 hourly modifiers and .50 pondage modifiers.

o Energy Sto~age ·- MAPS can model up to three energy storage plants.

o Operating Policy this data includes spinning reserve requirements, thermal commitment policy, maintenance and area protection requirements.

o customized Program Dim~nsions - Although the standard MAPS program is dimensioned as described above, the user can specify to us many of the dimensions in the program to suit their individual needs.

2.4 How MAPS Works

In order to enhance computational efficiency, MAPS is run in two parts. Part 1 sets up the original load models for each area and/or company ( s) and modifies the loads for firm purchases and sales and hydro. Part 2 schedules the storage devices and maintenance and performs the commitment and dispatch, recognizing the flow limitations between the areas. The logical program flow is shown in Figu~e 2-1. Part 1 and Part 2 will be discussed in greater detail in the next two sections.

I I I I I I I I

-----~

Jj

ll I '

I I ~ ~·;·

I ' .

Part I

Part I Input

• Set Up Company/Area Loads • Perform Firm Load Modif~cations ·

Part II • Perform Maintenance • Commitment • Energy Storage • Dispatch • System Production Cost • CompanyiArea Production Cost • Cost Reconstruction

Annual Output Int. Output

Part II Input

Figure 2.1 - MAPS Program Structure

2-4

I I I I

II .

CHAPTER 3

TECHNICAL .DESCRIPTION OF PART 1

3.1 Load Model

The Part 1 of MAPS is designed to make firm load modifications and set up the area and company loads for use in Part 2. In order to represent the area and company loads as accurately as possible, the annual load models are divided into 26 two week intervals. A typical week. is modeled within each interval, and 84 chronologi.cal hi-hourly load values are specified for the week. This allows the modeling of all seven days of the week in chronological order, without making any assumptions about typical weekdays or typical weekend days.

There are three options in specifying the input load data. ·First, the loads may be input separately for each area (see Part 1 variable LARCOM=l). For this option, each company's split of each area's load is specified in order- to construct the company loads . The second option, LARCOM=2, consists of inputting the loads separately for eaeh company. For this option, each area's split of each company's load is specified in order to construct the area loads. The third option, LARCOM=-1, consists of specifying both the company and area loads for each hour. In this mode, the program will check to ensure that the sum of the company loads equals the sum of the area loads for each hour.

Depending on the input option used, the program will first construct the remaining load data so that both the original company and area loads are available. At this point, any firm load modifications are made.

There are twa methods in which load modifications can be performed. The first method is to specify an hourly modification in the same form as the load model. For each hour of every interval the MW addition (sale) or subtraction (purchase) desired is input. Also specified is the area in which the modif•ication occurs , and how the modifier is divided among the companies. Modifiers of this type are termed run-of-river.

The second method of load modification is through the use of pondage modifiers.. With this method, the minimum capacity, maximum capacity and energy available for each pondage modifier is specified for each intei.~val. The program schedules the energy for each pondage modifier against either the area, company or pool load shapes to shave the peak loads. A graphical r-epresentation of run-of-river and pondage load modification is illustrated :tn Figure 3-1. If desired, the program will schedule the energy in the valleys rather than on the peaks.

3-1

•

~J

,, .

.\ ·r

'0

-::.:

I;

v 'l k·~i •:

i c ·,•

l

E I ~ .. ~

I I j ~' }

I I I l ~

I ' I .

-Ji ;

-~

~ 11 .

-~ ·' ~r ~ 4

~

~~i ., :._1

~.i

Rr, ~; "~

~ ~r _:L

~· j

'-~ :.:;:, .

As with the hourly modifiers, both the area of modification and the percent energy to eac~ company must be specified. With both methods, an energy charge in $/MWh for each modifier can be specified. The program accumulates a total dollar charge for each company f.or each hour. These charges are then passed on to the Part 2 program where they are accumulated with the company costs for the thermal units.

3.3 External Contracts

J,:n addition to the firm load modifiers, up to five modif'"iers may be .specified as ~conomy contracts in the Part 1. t2::,. pondage ~odifiers which ~ declared economy contractst an hourl~ s~hedule will be developed. The modifiers are then passed 4irectly to the Part 2 without changing the area or company loads.

In the Part 2, these modifiers have no impact on the PSH OFeration -or on the initial system t~conomic commitment and dispatch. After the initial dispatch in the Part 2, additional dispatches are then. performed with the contracts present to determine the gross system savings. This savings is then used in the pricing of the contract energy and in the determination of the net system savings. . For more information on external contracts see Sectton 4.7.

3.4 Inter-Company Contracts

Firm ~ontracts. between companies can also be modeled. With this option +100~ ownership of the modifier for the company that is selling the energy, and -100~ ownership for the company that is buying the energy is specified. In addition, no area of modification is specified. In this way, the accounting transaction takes place between. the companies, but no load has moved within the system.

3.5 Part 1 Dimensions

The program is cur-rently dimensioned to model 11 areas, nine companies, lO hourly m~iifiers and 50 pondage plants. If these limits are restrictive, please contact EUSED. Almost all program changes have come about as a direct result of customer requests.

Section 5 contains a ~esc!"iption of the various output swmnaries that are available frCJm MAPS Part 1. The different output sections are written to separate: file codes so that maximum flexibility is available. In addi'tion to the hard copy output, detailed hourly information can be ~7r1tten to either tape or disk for interfacing. with any user written programs. Followin3 the cutput descriptions are examples of the output for our three area sample system.

3-2

..

'~ ' .

I I I I I I I I 11 1' \

-1' -

J1

-~

I ~ li: ~j

• :J , _ _:;

'.,_

..I'"L.·:~

~~----~~--~~~--~~-----~--~~- ·~~·---~~~--~--------~~--~-""""' ···--·- ~""'·-~--~---~~---~--

MODIFICATION OF LOAD CYCLES TO REFLECT HYDRO AND CONTRACTUAL PURCHASES AND SALES

!000

' I

500

-fi-1

SAT. SUN. MON. TUES. WED. THURS.

---TIME

• CONTRACTUAL PURCHASES AND SALES CAN BE TREATED IN THE SAME

FASHION AS RUN-OF-RIVER HYDRO OR PONDAGE HYDRO.

e A CONTRACTUAL SALE TREATED AS PONDAGE HYDRO CAN BE

SCHEDULED EITHER AS AN ADDER TO .THE PEAK OR VALLEY LOADS

FRI.

I

1

' ...

"

. ·•

I

I I L

r.

•'

,' l

'i ., ; ·, :;

b

)

I '

,\ " \

,,

I ' f . ~

-; ,! i

-!

-l

CHAPTER 4

TECHNICAL DE:SCR!PT!ON OF PART 2

It is in Part 2 that the production simulation of the thermal units is performed. The cormni tment and dispatch of thermal units is done recognizing the limitations in energy transfer between areas. MAPS can then repeat com!nitment and dispatch for each of the companies on a single area basis, and allocate the savings due to joint operation to the member companies based on a split savings between the buyers and the sellers of energy. Separate multi-area dispatchen can be performed for determining the cost savings of any external contracts . MAPS Part 2 also has the ability to develop the maintenance schedule as well as the energy stora.ge schedule ..

4.1 Maintenance Scheduling

There are two ways in which to schedule maintenance when using the MAPS program. One way to schedule maintenance is to use the maintenance scheduling algorithms in MAPS Part 2. There is also a saparate maintenance scheduling program available with MAPS. When this option is used, the Part 2 reads in the maintenance schedules that are output from the maintenance scheduling program and bypasses the maintenance calculations in Part 2. The HAPS Part 2 variable NTAPE is used to indicate that a separate maintena..'lce schedule is to be read into the Part 2 . If the Part 2 maintenance scheduling is to be bypassed, then the variable NARRY(2) should be used. The maintenance schedule developed by the separate maintenance scheduling program can be overridden in the Part 2 by using the variables NMAINT, MSTAR, and MSTOS.

Thermal unit maintenance can be scheduled either manually or automatically on a bi-weekly basis. Manual maintenance may be specified for each unit for each of the 26 bi-weekly intervals in a year (variable MSTARA, MSTOPA). Multiple periods of manual maintenance may be scheduled for arty of the units by inputting MSTARA and MSTOPA on an interval basis. For automatic maintenance scheduling, the number of intervals that a unit is to be out on maintenance for a y~ar is specified. This is done by using a maintenance matrix which indicates the cyclic pattern of the unit maintenance (variable MARRY). If a unit is to have automatic maintenance, that unit is assigned to one of these patterns which can be up to six years in length (see variable MTKEYl). The ~osition within each pattern is also specified for the unit by using the variable MTKEY2. A unit that will not have any automatic maintenance will have a pattern of -1 and a starting posit.ion of 0. If a zero is used for the pattern, then the unit will be placed on one interval of maintenance. The program uses the number of intervals per year of maintenance information to schedule maintenance for the units on automatic maintenance. A system criterion of levelizing the amount of available reserves in each interval is employed. The unit with the

4-1

I I I I I

I I

I

largest capacity is scheduled first. Any combination of manual and automatic maintenance may be specified.

The maintenance scheduling p:t ... ogram has several added features over the scheduling routine that is in Part 2. In addition to the manual maintenance~ when the automatic maintenance is being scheduled the • program can levelize the amou:tt of reserves on an araa rather than system basis. Maintenance windows may be specified for each unit, indicating the allowable intervals in the year in which maintenance can be scheduled. Once the planned maintenance ha~ been scheduled, random maintenance may be scheduled. This option uses the units' forced outage rates to determine the number of intervals of random maintenance to schedule for each unit. These intervals are then scheduled using a random number generator. The maintenance schedules from the scheduling program can be overridden by input data to the Part 2.

4.2 Thermal Unit Data

Each thermal unit in MAPS is modeled with a minimum, continuous (CAPREO) and maximum power output rating (CAP), along with any number of intermediate power points. The total number of power points for the system must not exceed the value to which it is dimensioned. The number of power points may be divided in any manner among the individual units. The continuous and maximum ratings may be the same for a unit or there may be one or more power points t>etween them. A unit will be dispatched beyond its continuous rating only after -11 of the available units have been dispatched to their continuous ratings. For each unit, the .fuel input at minimum ( Ci~OLOD) and the incremental heat rate fer each delta section (variable BTU) rn~st be supplied.

Additional unit data includes fuel tyJ?e, fuel cost, minimum down time, fi4ed and variable operation and maintenance costs, maintenance requirements, and· forced outage rate. Figure 4-3 illustrates the capacity model that is csed by MAPS to represent thermal generating units.

To satisfy various area, plant, or company protection requirements a unit may be classified as a must-run unit. This is done by placing the variable MUSTRN equal to 1 for that unit. See the section on commitment for an explanation of must-run units.

Each unit is assigned to a specific geographical area (KNUMBR) and to a particular company (KCOMPY) or companies if the unit is owned jointly by mc:~.·e tliCi~ one company. A jointly owned unit will have a value of 0 for KCOMPY. The program will then look to the variable IDCOMY to see if that unit number is in that list. If it is, then NOCOMY is used to indicate the companies that own the unit indicated in IOCOMY, and COMYPR is the per unit value of the unit that is owned by that company.

4-2

~!' ,_'

I' t I

! !

ll !

I

The specification of the geographical location of each unit allows MAPS, when combined with the transmission data described in the next section, to perform a multi-area production cost simulation. The specification of company ownership for eac~ unit allows MAPS to calculate company cost from the pool dispatch, and when combined with company load models r~s can perform company dispatches. The results of the pool and company dispatches are used for the cost reconstruction calculations.

4.3 Transmission Data

A key capability of the MAPS program is its ability to accurately represent the limits to the free flow of energy within a utility system. The generation and loads are assigned to a maximum of 11 discrete areas, and up to 10 transmission interfaces can be defined. These interfaces can be arranged in any manner. Radial or looped networks can be modeled.

On each transmission interface the maximum transfer capability in MW is specified. For each transmission tie, the transfer capability in one direction may be different from the capability in the other direction. These limits may be changed on an interval basis. In addition, up to five full sets of limits can be input for the system each interval . Each set of limits is assigned hours for which it applies. Therefore, different transfer limlts can be u~ed ror on-peak, off-peak, weekend, or other load conditions to maximize the simulation accuracy.

4.4 Commitment

For each interval, the commitment process begins with the development of a priority list of those thermal units which are available to serve the load. A thermal unit is ·available if it is installed on the system and is not on maintenance.

The priority list is an ordered list of the thermal units which indicates the sequence in which the units will be committed to serve the load, with the first unit in the list being commj.tted first. The priority list is based on the full load cost of each available unit. The full \cad cost is the $/MWh cost of operating the unit at its continuous rating and is calculated from the fuel cost and heat rate .. Variable operation and maintenance costs may also be included in full load cost if the variable KVONM is set to 1. A penalty factor may be used to adjust the priority order. The penalty factor (variable PF) is a per unit multiplier of the fuel portion of the full load cost. If variable operation and maintenance is included in the full load cos·t, then that portion will not be changed by the penalty factor.

There are two methods in which th~ priority list may be determined {see variable KOMIT). In the first method, the position of each unit in the list is determined by the unit's minimum down time (variable LAWOPH)

4-3

o I

I

I' • i. I'

Jl

,,

I

and it's full load cost as described above. The order•of the priority list is determined first by sorting the units by their minimum down times and then their full load costs. The units with the highest minimum down times have the · highest priority. For units with the same minimum down times, the ones with the lowest full load costs will have a higher priority. In the second method, the priority : .. s determined only by the full load costs of the units. When using th~.s method, m:.Jtimum down times are not consiciered in determining the ordnr of priority but they are recognized when doing the commitment.

If a unit is classified as a must-run unit OtuSTRN), then the minimum power section of that unit is taken first wheiJ comparing the sum of the continuous ratings against the load. The re:naining portion of that unit, i.e., continuous rating minus minimum ratLng, is put in the proper position in the priority list. Only the mi:.1imums of must-run units must be compared against the load. The remair.ing portion of the unit will be turned on if it is economic and if it dues not violate the tie limitations by being turned on.

There are three commitment options in the MAPS program (see variable ICOFIX). In all three options, the preliminary commitment is done based on the following two criterion. The sum of the co: . .ltinuour; ratings of the available units must be greater than or equal to the load and the sum of the maximum ratings of the units must be grea1;er than or equal to the load plus the spinning reserve requirement. T.1e spinning reserve may be specified for the pool, company, and area ( ~ ee 'Jariables SPINA, JSPIN, COSR, ARSR). The commitment is done for each program hour, i.e., 84 times each interval. Since th@ commitment is devt~loped each hour, no assumptions are made about "zones" of constant commitment. This commitment is ·then modified to re'!ognize the cycling capabilities of the units by obse~ving their minimum down time requirements. The commitment of the thermal units is illustrated by Figure 4-2~ Each commitment option will be described and how it relates to thu different spinning reserve requirements.

In the first commitment option (ICOFIX = 0}, a separate commitment is performed for the pool &nd the company. The pooJ. commitment is done using the continuous ratings of the units, the area loads and recognizing the tie litnitations between the areas. If it is desired to meet the spinning reserve on an area basis (variat·le JSPIN) the area spinning reserve (SPINA times ARSR) is checked and additional units will be turned on until the sum of the maximum ratings of the units in an area is greater than or equal to the area load plus the spinning reserve requirement for that area. Then, because the sum oi' the area spinning reserves for the areas does not have to equal the pool spinning reserve, additional units are turned on if necessary to make s.iJre that the sum of all of the committed units are greater than or equa:. to the pool load plus the pool spinning reserve r~quirement.

l

4-4

I . l

' '

• 'V. I

'~' I

.I

u

' ' ~

' ,.

" ' . I

11

After doing the commitment and dispatch for the pool, the commitment is then done for the companies. In doing the company commitment, the sum of the continuous ratings must be greater than or equal to the company load and the sum of the maxiDlum ratings of the units must be greater than or equal to the company load plus the company spinning reserve.

When using the second and third commitment options, ICOFIX = 1 and 2, respectively, separate c9mmi tments are obtatned for each option. However, the same commitment will be used for both the pool and ~ompany dispatches.

The second commitment option (ICOFIX = 1) first does a pool commi.tment to be sure that the sum of the continuous ratings of the thermal units is greater than or equal to the load as was done in the first commitment option. Then additional units, if any, are turned on to make sure that this ~ommi tment also satisfies the company cornmi tment. The company commitment is done when the company spinning reserve is zero or greater (see variable COSR) u Wh~ui the company spinning ~eserve is negative, then that company is not checked to see if it meets the commitment criteria and, therefore, that .company • s spinning reserve is not met . The units commit ted under the pool ar-e surmned for each company taking into account the jointly O'-'"Iled units. The sum CJf the cont1.nuous ratings must be greater than or equal to the company load and the sum of the maximu..'O ratings of the units must be greater than or equal to the company load plus the company spinning reserve. Next the user has the opt!lon to check the area spinning reserve (see JSPI:tJ) • Finally, the pool spinning reserve is then checked to be sure that the sum of all of the continu.:>us ratings of the committed units are greater than or equal to the pool lc)ad plus the pool spinning reserve.

l'.he final commitment option is for the company commitment to be used for both the pool and company dispatches . Thts company commitment is done in the same manner as was done in the first conmtitment option, that is, the sum of the continuous ratings must be greater than or equal to the company load aml the su.m of the maximum ratings of tt~e units must be greater than or equal to the company load plus the company spinning reserve. The spinning reserve fer the pool is the:n. checked and additional units are turned on if necessary.

4.5 Energy Storage

After the commitment has been developed, the program schedules any energy storage devices, ~;uch as pumped storage hydro, that may be present. This co(mni tment is called the pumped storage commitment. The commitment that is used fo·r scheduling the storage devices is the pool comrnitment (multi-ar-ea · commitme:nt) since 'the stora.~e devices are scheduled against the pool load. The scheduling of the storage devices begins with using the commitment for each hour to construct a cost curve. The cost curve is a function of the thermal characteristics of

4-5

I

\)

..

"

")

"

0 .. I ~ ,.

~)

' ~..Ji " ''[

.• ,. (f,~ ;:') ,.

/''

. a·

''i'

~("' ., /.1

•I ;<;

"

-:, ~

.. '

I I I I I I a ~ • = a I I I I I -~ J

J)

IJ

Mt

I •... :J

~J

the commit ted units and the prior-ity list. The cost curve is used to determine the most economic operation of the storage devices on the system.

The program will allow up to three separate storage plants to be modeled on the systgrn, and each plant is assigned to a specifin t:trfHL

The storage generation and pumping is scheduled to levelize the pool loads as much as is economically feasible. This scheduling is done against the chronological weekly load model. Energy reservoir levels are monitored to avoid any physical violations of the energy storage. In Figure 4-3, the relations hip between the cost function, load shape, and the storage reservoir is shown using a simplified example .

For illustration purposes, 1 t will be assumed that one pumped storage plant is present. An economic operating schedUle must be developed for the plant.. The following is known:

o The sequential load model and spinning reserve requi~ements for the week. These take into consideration all load modifiers such as pondage hydro, run-of-ri~Ter hydro, and. firm contractual purchases and sales.

o All thermal units available for scheduling during the week. In addition a priority order of these therm~.l units would be known.

o Characteristics of the pwnped storage plant such as the maximum generation, maximum pump rating, reservoir limits, refill policy, cycle efficiency, and spinning reserve credit.

The first step taken by the program would be to development of the hour-by-hour thermal unit cormnitment schedule that would result assuming that there was no pumped storage plant presen'".. Front this weekly commitment schedule, the program determines the maximum and minimum thermal unit commitroent that was make during the week.

The program then develops a thermal cost function relating the hourly thermal costs ($/hr) to thermal power output (MW). The manner in which this cost curve is built up is shown by Figure 4-4.

Note that this curve is constructed so as to be convex. Convexity in the regions 4-7 is assured by the use of a priority list where the unit rank is determined on the basis of cost ($/MWh) at the continuous ratings of the units. Convexity in the regions 1-4 is obtained by allowing units to go from minimum output to continuous in economic order. That is, the unit tha~ has the lowest $/MWh cost will be loaded first.

It is conceivable, but not likely, that th~ slope from say point 3 to point 4 to be in excess of the slope frcm point 4 to point 5. In this event, special program logic is employed to restore convexity. It

4-6

\I

I

I I I I

Cl

I I I 11

11

IJ

is essential to the program that a convex function be obtained as a function that is non-convex can cause a premature termination in the development of the pumped storage hydro operating sched,.i.l~.

There are at least two cost functions developed for each week; one for the weekday period and one for the weekend period. The functions are necessary as the number of units committed in the valley hours tend to be different on the weekdays from the weekends and this difference in cormni tment does cause a cost difference from the same system thermal load. Actually, the program has the capability of developing cost functions for several time periods during the week should this pr9ve necessary in the course of the schedule development.

Figure 4-5 shows how the cost function of the thermal generation is combined with the load cycles of the system to arrive at a feasible and economic pumped storage hydro schedule. For purposes of illustration, a much simplified representation of tue weekly load cycle has been used. The procedure used by the program in developing a pumped storage schedule is as follows:

0

0

0

0

0

The program begins by assuming no pumped storage operation. This in effect says that the load SHAVE line rests atop the highest load and the load FILL line rests atop the lowest valley. In Figure 4-5, this condition is shown by SHAVE(O) and FILL{O). As no operation is assumed, reservoir storage is equal to the maximum reservoir storage for the entire week, and this is shown as STORAGE{O) in Figure 4-5.

An increment of pumped storage generation is called for by lowering the load SHAVE line from SHAVE(O) to SP~VE(l). Referring to the Cost Function, it can be seen that a savings of 5, 400 $/hr is achieved by this action ..

The water used by this generating action must be returned to the reservoir in order to meet the requirement that ending storage be equal to beginning storage. To do this, pumping action is called for by raising the load FILL line from FILL(O) to FILL(l). A cost increase of 1,800 $/hr results.

Note that SHAVE(O) - SHAVE(l) ::: 1000 MW, whereas FILL(l) -FILL(O) = 1500 MW, although it has been stated that the same amount of water has been returned to the reservoir as was taken from it. This is due to the pumped storage plant having a cycle efficiency. For every two MWh of generation, th.ree MWh of pumping energy must be expended to return the water to the upper reservoir. The user may specify the cycle efficiency for each pumped storage plant by using the Part 2 variable CYCLE.

Despite the two for three ratio noted above, moving from (0) to ( l) has resulted in a net savings of 3, 600 $/hr ( 5, 400 $/hr -

4-7

~ ·' ..

·,-

;

f I

I­I I I

I I IJ

I IJ

I

0

0

l, 800 $/hr) . This step is proven to be economic and will be taken if it is feasible.

Feasibility is tested by first checking to see that the generation and pumping m&gawatts called for are within the capability of the plant. A second feasibility check is made to be sure that the proposed schedule change does not cause a violation of the storage reservoir. The storage level is checked for each hour to be sure that it equals or exceeds the preset minimum storage level and is less than or equal to the p~eset maximum storage level. If no constraint violation occurs, the change to the schedule will be made.

The process described in the preceeding steps in repeated over and over again until an economic loss is encountered. When this occurs, the SHAVE and FILL lines are moved back until ~n economic gain is encountered again.

Figure 4-3 is very siEilar to Figure 4-5 except that a more realistic weekly load cycle has been used. For convenience in illustrating the cycle, the week has been shown to begin at midnight Friday rather than early Monday morning, the normal starting point of the program. This figure clearly shows that although the schedule development stops when incremental costs exceed incremental savings, the overall ~chedule does produce a total net savings.

The data required to describe a pumped storage plant is not at all extensive. This is possible because of some of the assumptions made concerning plant representation and operation. The following is a list of these assumptions:

0

0

0

No attempt is made to represent the effects of head variation.

A constant cycle efficiency is used and all losses are taken in the pumping mode.

Pumping and generation can be done continuously over the range of their respective capabilities.

The schedule shown in Figure 4-3 starts with a full reservoir on Monday morning and returns to a full reservoir on the following Monday morning with no intervening storage constraints other than those of reservoir minimum and maximum. This permits the development of a weekly cycle of operation. This cycle can evolve into several daily cycles if this should prove economic, but it is important to note that daily cycles are not forced by this approach, but rather develop if the economics dictate.

Generally, a weekly cycle does much of the refilling of the reservoir on the weekend due to the weekly load shape. This causes the

4-8

a 'lc

a I

(;'

a i'

a a a c a

.

= I

~ ' ,,

I (·.

i'' I il

" I I I

f·

I -/ D

system to begin successive Meekdays with less and less storage. If this declining storage condition is of concern to the user, the program has the option to require that weekday generation occur only if the water used can be economically returned to the reservoir that same night. This daily refill causes the system to enter each weekda:l with a full reserv.oir.

After the pumped storage has been scheduled, the loads are modified to the reflect the action of the plant. At the users option, output may be obtained showing this modification and the resultant system load.

A spinning reserve credit for pumped storage hydro plants is either set to zero or can be calculated by either of two methods. These are shown as follows:

Condition Spinning Reserve Credit Mode I Mode !I

Generating G maximum G actual G maximum - G ~ctual

Inactive G maximum G maximum

Pumping G maximum P actual

Note: G = generate P = pump

If more than one pumped storage hydro plant is present, two approaches can be used. If the plants are proportionate in terms of power and. storage and have the same cycle efficiency, the plants may be combined into a single equivalent plant. When doing a multi-area analysis, the additional criteria of the pumped storage plants being in the same area is imposed. If the conditions just noted cannot be met, the plants can be considered in series. In this case, the order of the plants as specified as data to the program, is the order that they will be scheduled.

Once all storage devices have been scheduled, the commitment process is repeated using the loads which have been modified by the operation of the storage plants. This commitment is what is called the final commitment. The same options a~e used in the pumped storage commitment as they are in the final commitment.

If the commitment and dispatch is to be performed by derating the unit • s capacity by the forced outage rates, (see variable KRFOD) then both the pumped storage and final commitments will use derated capa.ci ties. If a stochastic dispatch is to be performed, {see variable KRFO) then the pumped storage cormnitment is performed using derated

4-9

'• ,.

1\ \

c c

I I I I

I ~ I ·~ ~

D

. I

capacities. However, the actual capacities are used when doing the final conuni tment. If no pumped storage is present, then the actual capacities are used.

4.6 Dispatch

Once storage has been accounted for and the final commitment has been developed, the thermal units are dispatched on an hour by hour basis to meet the loads. The dispatch begins with the loading of all committed units to their minimum operating points. Units designated as must run unit (MUSTRN) are loaded to their minimum operation points first. The delta sections or the committed units are then loaded in the order of increasing costs until all the loads have been served. A penalty factor may be used to adjust the priority order. The penalty factor (variable PF) is a per unit multiplier of the fuel portion of the full load cost. If variable operation and maintenance is included in the full load cost, then that portion will not be changed by the penalty factor. The same penalty factor is used for the commitment and dispatch. An efficient linear programming algorithm is used to dispatch the units in the most economical manner without violating any of the tie constraints. For a detailed explanation or the linear programming method used by MAPS see A. K. Adamson, J. F. Kenney, A. L. Desell, L.L. Garver, "Inclusion of Inter-Area Transmission in Production Cost Simulation," IEEE Vol. PAS-97, No. 5, September/Oct.ober 1978, pp. 1481-1488 which can be made available upon request. Each hour is dispatched separately so that load diversity, commitment and tie capacity changes can be treated accurately.

The thermal unit dispatch may be performed either. deterministically or probabilistically. Forced outages of the thermal units can be modeled in MAPS in three ways: deterministically using deration (variable KRFOD), stochastically (variable KRFO), and "random maintenance." The "random maintenance" is determined in the separate maintenance scheduling program. Thermal forced outages can be input for the whole unit (OUTRT) or input for each power section or each unit (POUTRT). The variable MSTATE is used to determine which forced outage rate variable to use for the calculations. If MSTATE is 1, then use partial forced outage rates, i.e., do multi-state calculations.

The cost of production is determined by summing the cost or each of the sections of the thermal units that were dispatch. The actual costs are used and not the costs adjusted by the penalty factors.

4.7 External Contracts

The MAPS program has tht~ capability or modeling economy contracts from outside the utility system being studied. These contracts can be modeled in two ways.

First, the amount of contract energy available each hour can be specified, as described in Section 3-3. Full multi-area dispatches are

4-10

~'; \

-:<' }

-~·\.,_

\~~)l ~1

t • '< }'

' '

l t I,

I I z

I

~

~

f ~-

r r r r r t

r

f I

then perf"ormed with and then without the contract energy, as shown in Figure 4-6. The cost difference between the first and second dispatches is then used to determine the price paid for the contract energy and the net system savings that is a result of the contract. Figure 4-7 illustrates the relationship between the dispat~hes with and without the contract present and the savings which result from the contract.

The :;;econd way of modeling external contracts is to specify the maximum ciJilount of contract energy available and. a "trading" cost. As with the first type of contrac:t discussed above, two dispatches are performed~ one with and one without the contract. For this second type of contract, in the dispatch with the contract present, the contract will displace only thermal units whose incremental costs exceed the "trading" cost of the contract. The difference between the first and second dispatches is used to determine the price paid for the contract energy and the amount of contract energy purchased.

If more than one contract is present, the order in which they are evaluated is specified by input data and the contract savings are determined sequentially. In all the dispatches discussed above, full recognition is maoe of transfer limitations between the area the contract is available in and a.ll other areas in the utility system being studied. If cost reconstruction is being performed, different options are available which dictate the manner in which the contract savings are distributed to member companies.

4.8 Cost Reconstruction

Often the primary results or a production simulation analysis are the dete~ination of the total system savings which result from integrated operation. Sometimes it is desireable to detennine how the savings are shared by the member companies. The cost reconstruction option in MAPS calculates these savings. There are two types of dispatches performed in MAPS as shown in Figure 4-8. First, a multi-area dispatch is performed in order to determine the savings. The MAPS program then repeats the commitment and dispatch process for each of the companies on an isolated "own load" basis. The commitment options in the MAPS program are described in Section 4-4. These two dispatches are then compared to the system dispatch on an hourly baGis to determine the companies which are selling powe~ and their incremental cost, and the companies which are buying power and their decremental cost. A full cost reconstruction is then p!:;rformed, with the savings being split between the companies which are buying e.na the ones which are selling.

Figure 4-9 exemplifies four companies that partake in a split savings cost reconstruction. Two companies, A and B, were found to be sellers. For companies A and B, the generation in the multi-area dispatch was greater by 50 MWh for that hour than it was for the "own

4-11

! ' r.

•

, . .

l

r r r r r r

load" dispatch. The change in cost divided by the change in MWh for company A resulted in an incremental cost. of $20/MWh. Company B has an incremental cost of $25/MWh. The average incremental or sellers cost was calculated to be $22.50/MWh = {50 ~~ * $20/MWh + 50 MWh * $25/MWh) 1 (50 MWh + 50 MWn). Companies C and D generated less in the multi-area dispatch than in the "own load" dispatch and are therefore called buyers. Company C bought 25 MWh at $30/MWh and company D bought 75 MWh at $35/MWh. The average decremental or buyers cost was calculated to be $33.75/MWh.

The sav~ .. 11gs from joint company operation can be seen on Figure 4~10. Company A's savings results from· half the difference between the average buyers cost and company A • s selling cost times the amount of energy transacted. Company B's savings is calculated in a similar manner. The savings for company C results from half the difference between company ct s buyers cost and the average selling cost times the energy transacted. The other buyer's savings, company D's savings, is calculated in a like manner. The sum of the savings for the sellers is equal to the sum of the savings for the buyers . A per unit of each company's savings can be placed in what might be called a transmission fund by using the variable TRPCT.

4.9 Output

A number of output summaries are available on both an interval and an annual basis. In addition to printing the dispatch of the units under pool operation, area summaries can be obtained. MAPS also provides company summaries, and interval and · annual summaries of the savings resulting from pooled operations. Summaries are also available showing the energy transfers between areas and loading curves for the ties.

An output tape can also be created which contains the hour by hour marginal cost for each area and company. This tape can then be input to an auxiliary marginal cost program to develop on peak, off peak, monthly, and seasonal system marginal cost.

Sections 5 and 6 contain sample output from the HAPS program along with descriptions of the different output sections. Section 5 contains the Part 1 output and Section 6 contains the Part 2 output.

4-12

•·

D a

\\; . g ,. D

~ . ' ~

';' l· ~ il ~

'I J..i·

ll r ~ . 1 ' 1

r r ~ f;

•.. ·j , ,··/. J

9 '

· ... r r r r '

r r

COST DATA

FUEL COST O&M COST START UP FUEL INVENTORY

PERrORMANCE DATA

CAPACITY MODEL

$/MBTU $/KW~ S/MWH~ S/HR $/COLD START~ HRS. TILL COLD START $/YEAR

INPUT - OUTPUT INCREMENTAL

HEAT RATE

FUEL INPUT i (MBTU) .

HR I FMIN+

PMIN PMAX

P-QWB~ UU1Pil1 C;M~)r · •· · ... ··

OTHER

MAINTENANCE SULFUR CONTENT FUEL TYPES AREA MINIMUM DOWN TIME

Figure 4.1

4-13

-KWH

POWER~

• ·~--.~~., .... ~~~--.~~;.

,',1<

I' r

a D

a a·

D . '

~·

~

~

r r J

1 ,,.

~ '( r f

f r ~ ·;

r r

r f l

..

=

COM M ITT MEN T OF THE R M A L UN ITS

17 17 !17 1 '17 17 L. • ..J ~

15 15

12 12 12!

~ I ..,......,. • _____ _J

LOADS TO BE SERVED

_, ___ _ PREliMINARY OR MINIMAl ~OMMITTMENT FINAL. OR REVIS£0. COMMiTTMENT

IN THE PRELIMINARY COMMITTMENT N

L PcONTn ~ LOAD n=l

N

L PMAX,. ~ LOAD+ SPINNING RESERVE ft&l

IN T HE FIN A L C 0 M M I T T MEN T, IN A D 0 I T I 0 N T 0 T H E A 80 V E • N IN I M U M DOWN TJ ME R U L E S A R E M E T.

Figure 4.2

4-14

(]

··~ '

~";I

I l

t!

J J

J ... J

~

r l

J

t THERMAL

3500 COST

C$1 HR)

I

0

-PUMPED STORAGE-

THE DEVELOPMENT OF THE THERMAL COST FUNCTION

. PRIORITY IDENTITY P MIN $1 HR AT P CONT .

ORDER (MW) P MIN ( MW) ..

I A 100 300.00 300

2 8 100 330.00 300

3 c 100 360-00 300

4 D 25 X 100

5 E 25 X 100

6 F 25 X 100

MIN I M U M C 0 M M IT M EN T FOR P E R I 0 D = 3 U N I T S

MAXIMUM COMMITMENT FOR PERIOD = 6 UNITS

POINT p $/HR COMMENT

<D 300 990.00 A, B, AND CAT MIN. LOAD

® 500 1530.00 8 GOES TO MAX. LOAD

® 700 2100.00 A II II II II

@) 900 2700.00 c II II II II

® 1000 3120.00 D ADDED AT PCONT

® 1100 3600.00 E II " II

(i) 1200 4200.00 F II H II

200 400 600 800 1000 --THERMAL MW ----1.,...,

Figure 4.3

4-15

$1 HR AT

p CONT.

870.00

870.00

960.00

420.00

480.00

600.00

1200

.~~--... ~----~~~~#~ ... ~ ·. . ,.+&; • ,_.~~~~~---·~- -~.-~ ij~~ ~~,:~.; •• ;::·: .. ~-;~7'~~~v:as;csr :~1. ·~·

(l

!} I

~i

(

I

f (f)

a: ::> 0 :I:

A SIMPLIFIED ILLUSTRATION or: PUMPED STORAGE HYDRO SCHEDULING

COST FUNCTION

GAIN = 5,400 $1 HR

1----+--~~to,soo $1HR

NET GAIN= 5,400 -I, 800 = 3,600 $/HR

A-·--+---+-....., 6, 6 00 $ /HR LOSS = 1,800 $/HR

0 2000 4000 6000

--+-MEGAWATTS+--

2000 4000 6000 0 0-+-----+--+-..-+-..----+-.--+---+-

24

48

72

96

120

144

WEEKLY

LOAD

CYCLE

-0 -...J ...J

---...J ...J - -LL. LL.

--LLJ > <t :I: en

168~-------------~--

4-16

-0 -LLI > <t :I: (f)

-~RESERVOIR STORAGE

0 -r-r--------r--

24 1 STORAGE { 0)

V> 72 a: :::> 0 :I: 96

+ 120

144

STORAGE (I)~

/

( I ~

/

' ' ' RESERVOIR STORAGE

168~----------------+

MIN. MAX. STORAGE STORAGE

•j '~

1] ' 1"~--..-'(----~-, -· .... ~u-~.._.,..,............_+..,..~--~:1'.' ~ ...... "£'~~~~~~~~.I

Y1 ,- * ·- ---=· • "'''· -•>·~~~·"'~.:,.~\.;~~ .j

J

0 a 0 a a 0

p~

a ~

f I r ~-·

J J

'.-..

THE DEVELOPMENT OF A WEEKLY PUMPED STORAGE SCHEDULE

l $/HR.

CJ) t­C/)

0 u :E. w ...... CJ) > '"

SAT.

FRI.

1 A A_A_ COST SAVINGS

I

f j=y' COSTINCREASES ---+-- -~ • -, .- ' __ ,__ _.1_ ..

-MW--+

Figure 4.6

4-17

.. 4

-

,]

. I I

• fit~ ,, w•

. . "' . . . . . . . . . . . " ,, . :' . .-. .. :·:., . . . . @,. ,;.· v. :.~~J;;·~:' -··~·····---·~~~--MllliiiiiTII!!liiiiWilt1 i!!!f •• IJiblii'IJ.$•N71J'¢!r lifftz i\-.,. T 1 e'lno , ·""'~-~-~--~------- ~~--·-'...,j"'

-··---·- ~

. ' ·-~-~... ..,._, ~ - -- fr"1liiil ..-.. ,.._... 1lr"'"""1' ~ r--111 ..--. ~ ~ t-11 :I . . - . -. . """"" .,,,,.. """"' - "'·-" ,__, """""' ._.. ....._.. ..........

'r lf • ! ·~. . I MULTI-AREA MUL TI·AREA ~ / DISPATCH DISPATCH k I I l I I

I

~ I

J-a (X)

WITHOUT EXTERNAL PURCHASE

• TOTAL COSTS WITHOUT EXTERNAL PURCHASE

WITH EXTERNAL PURCHASE

• TOTAL COSTS

EXTERNAL PURCHASE

WITH EXTERNAL PURCHASE

• COST OF EXTERNAL PURCHASE • ALLOCATE SAVINGS FROM EXTERNAL PURCHASE

Figure 4.5

f l

I I !

I

!

f f

,,.

I

f '··~

- _'{

' ,i ~t'f II

·~1 ·~~(·:::!~~ >;;a·;~~;

,:-:;"'~ : -~:~;~~~-~

-

I i ! l

::--d

i 1

. 1

' f

I l t J

. 1 ! i

I 1

- -~ )j

, I "'

~ e c ~ • • • '• • •, " • . . •• • • • • % ''• ·-;..;t, ~-~-~~~~~~~~~~-·~~~-~--~•••-•••••'m~Rs_,_,~-•••• ~~-~-----~~ I t:::l = ca .. - - - ._ ~ J;;)l c:l t*ll CJ t:J LJ Cl t:::J t:::J Cl r} ' ' f . I ' ·~. : .. ! r . I r .. ;

1 MAPS OPTIONAL EXTERNAL t I CONTRACT REPRESENTATION I. I ' I ! ' . I I r i I I W/0 I

: EXTERNAL I

~ I ._. HOURLY

SYSTEM PRODUCTION

COST

PURCHASE FINAL I •- u - --• SYSTEM j··

WITH E·XTERNAL PURCHASE

COST 1

- , I· I· .

t..O

s

r.·.·.·.·.·.·.·.··:·:·:·• } : .,: .. :-:· :· :::: ~===· :-=·: I .... ·.· ........ I :::::::: .; . ~.: •,: ·:: :::::; ...................... ·. ~ · .. ~--·.·.·. ·.· .. ·.-.... · .... , '!'7:" ............. .

PURCHASE COSTS

I

.

PURCHASE ENERGY COSTS = PERCENTAGE OF DISPLACED COSTS 0

C)

Figure 4.7

~""',.~;(· "~· ... 4 pa-~

~ .. . ~

-.• ·- " . ~ .. . • .. - . . . .. - -~ = ··i·. . ••. ,

~

. ,, ':I i t:::::l lC:ZI E1ll - ... - - - i'ilfi1il l:li ~;:· ]I c. 1::1 Cl [:::J Cl Cl r:::::J 0 ': .

\~_;.':::-,

I I ·~ I DISPATCH DISPATCH f

I 8 0 I !

! I

(! -::~

., !1

~d ~:. I ·- .... '. .,...

...p. I

N 0

0 l

• TOTAL COSTS • TOTAL COSTS • AREACOSTS • COMPANY COSTS • COMPANY COSTS

~-..... .,....-~----.,.,.___ ----

• DETERMINE SAVINGS DUE TO POOL OPERATION

• ALLOCATE SAVINGS THRU HOURLY COST RECONSTRUCTION

• ADJUSTED COMPANY COST

Figure 4.8

:\

r I

,,

_,_.

}l··t,~ ·-··l/. -~ "!~- ;,.~"'.' ·::~. ' .,.., ·'J' k \'>·-s ..

~·~- Wi

• '0 ,, , .. ,. • ··"<C

-=----MHm...,._,. ----.. ··-·-···•·~· . . ---~~-~~--·-·~---C..-..--··--~--~-:~-~ .. c:::J CJ !CJI Ell - llll'll .. JIQI l:ll CJ [_J Ll c:J CJ c:J C1 E:l L:J C l ~ . • .,

I

iillliL.-~ ··- '"'

~

• N ......

I , l 0

l 1: I

MAPS COST RECONSTRUCTION I . SELLERS

COMPANY A-50MWH AT $20/MWH

COMPANY B-50MWH AT $25/MWH

AVERAGE INCREMENTAL . (SELLERS) COST=

2250 = $22.50/MWH 100

/

Figure 4.9

BUYERS

COMPANY C-25MWH AT $30/MWH

COMPANY D-75MWH AT $35/MWH

AVERAGE DECREMENTAL (BUYERS) COST=

3375 = $33.75/MWH 100

l

r I 1 l

) ! ,, l

I

I

l r

I (

I'

~ I

-.--.~-----=. ---:c,---;;;-:;·.---:-;-.;_~.. --. ------------,----- ----,--,---,-----~-- -------.--.. - ------:c-::---;-·::;----,..-------:::--~--,----c:-·;<~:;; ·.- ~~ -,.~' --~ -~ --· _, ___ "" ,,. ~- ·-~ ·+-• -;-·

~ :j

j,_·

1 ________ ...... ~·····"--------"-:.:~·,~t~ ..... ,.,_~_ ~-~~~~~<..M'J!I:a'llfJIII!~olidMl81llll•A• PtM·~_-· -1!Z'Illi\<,........_ n e ...... v ·.. ~ . -------~ , ·~·· .......... ~.___..~. --~---.-:--" ·· .. ·.· ' ' h · ! t __ _j CJ ell B:JI 1111 .. .. E:.il 1:.:1 Cl c=J L-::::1 CJ l_J CJ E:l Cl E3 CJ (_·· _._>: '_ I ~ , l ! •;.

+

I l I I j i

t ~ 1

.. ,

t I !

I I

I l

l

I I I : 1

I

I . l

..J::> I

N N

COMPANY

A

B

c

D

MAPS COST RECONSTRUCTION

33.75- 20

2

33.75- 25

2

SAVINGS

* 50= 343.75

* 50= 218c75

( 30 - 22.50 \ * 25 = 93.75 \ 2 ,

35 - 22.50 \ * 75 = 468.75 I

2

Figure 4.10

\

r I

' i L '1 ' • •e

. " . .. . : ' ' ..• --~-~--~-·------~·

" . . ~ .

....:c~__. ............ ·c..- .•• - -----~ • .._ ····-- v , ...... ~. -. _,. ,,,.,.._ ~----- ; .,. __ , ·--k•

$562.50

$562.50

l .

I

l·

I

\

['

I

I

<\

r:;,

"\.:::

".(1

•'Y ·~

...... -_.q ... ~ •

,.-1 " l f

I

L

• - ~·· Q ~

........ -- ._ ' .... ... - - M - ~-- ·~· - - - - - . - • Ill

GENERAL ELECTRIC COMPANY

ELECTRIC UTILITY SYSTEMS -ENGINEERING DEPARTM.ENT

1 RIVER ROAD BLDG. 2, ROOM 600

SCHENECTADY, NEW YORK 12345

GENERAL fj ELECTRIC

•. . - ' iii"] -1

..:

....... ...,... >•--,._. . ., ~· ··------......... ~·-···· ..•..• --- -·~ ~.-..·-~·····~~.-..,-·-.-~---··

_j_