A reliability assessment of the western canada grid

Thompson, Non Member Power Resource Planning Department Manitoba Hydro, Winnipeg, Canada Abstract This paper illustrates the reliability assessment methods used in the interconnection b

A RELIABILITY ASSESSMENT OF THE WESTERN CANADA GRID ©

P.R.S Kuruganty, Member IEEE

P.R Thompson, Non Member

Power Resource Planning Department

Manitoba Hydro, Winnipeg, Canada

Abstract

This paper illustrates the reliability assessment

methods used in the interconnection benefit evaluation

of the proposed Western Canada Grid interconnecting the

three Canadian provinces of Alberta, Saskatchewan, and

Manitoba A method of choosing compatible risk indices

for the three systems used in the Load Carrying

Capability (LCC) evaluation is presented These

standard risk indices are obtained from the system

studies without considering the grid interconnection

A study year is selected during which each system has a

risk level that approaches the design limit The

reliability benefits with the grid are obtained as the

improvement in the LCC's of the three systems with the

grid compared to those without the grid for the 1999

study year The effects of increasing the capacity of

the grid tap at Saskatoon on the reliability

performance of the Saskatchewan and Manitoba systems

are investigated Finally, the areas which need

further investigation in the reliability assessment of

the Western Canada Grid are identified

1 INTRODUCTION

Power system interconnection is an effective means

of improving the overall system reliability The

diversity existing between the interconnected systems

in regard to the loads and generating equipment forced

outages will allow each system to operate on less

reserve than would normally be required for isolated

system operation The actual benefits associated with

the interconnection depend upon the operating reserves

in the individual systems, capacities and outage rates

of tie lines as well as the type of agreements existing

between the systems regarding reserve sharing and emer-

gency assistance The application of probabilistic

methods in the interconnected system reliability

evaluation provides an analytical approach to include

these factors in the adequacy assessment procedure

There are several methods available for_ the

reliability evaluation of interconnected systems(1!-4),

The choice of the method used in a given study depends

upon the complexity of interconnection, computational

requirements and the type of results sought This

paper illustrates the application of the capacity

assistance probability method? to the reliability

assessment of the proposed Western Canada Grid system

83 WM 048-6 A paper recommended and approved

by the IEEE Power System Engineering Committee of

the IEEE Power Engineering Society for presenta-

tion at the IEEE/PES 1983 Winter Meeting, New

York, New York, January 30-February 4, 1983

Manuscript submitted July 19, 1982; made avail-

able for printing November 15, 1982

system is typically thermal

R Billinton, Fellow IEEE

‘Power Systems Research Group University of Saskatchewan Saskatoon, Canada

2 THE RISK CRITERION

The objective of the study reported in this paper

is to evaluate the reliability benefits associated with the interconnection of the three Western Canadian Provinces of Alberta, Saskatchewan and Manitoba to form the Western Canada Grid The basic steady-state reliability measure used in the study is the Loss of Load Expectation (LOLE) index The general procedure

to evaluate this index and the salient features of a computer program for reliability studies developed at the Manitoba Hydro are available on request The basic assumptions used in the study are as follows:

- The load models used in the study are based on the average daily peak loads for the months of December and January These are the months during which the system annual peak loads usually

- The load forecast uncertainty is assumed to be represented by a normal distribution with the system peak load as the mean value and 44 co-

assumed for the forecast demands of all the

- The peak loads used for all the systems are assumed to be the simultaneous peak demands and

no allowance is made for diversity between the various system loads In interconnected system

studies, the assumption is made that the maximum

assistance available from a system is the diff-

peak load, limited by the capability of the inter

system ties These assumptions result in conservative estimates of the system adequacy There are two alternative ways of evaluating the benefits arising from the grid interconnection; they

a) The improvement in the LCC for each system to maintain the same risk index (LOLE) without the

b) The reduction in the risk indices of the systems

if their peak loads are maintained at the same value without the- grid

In this study, the above two alternatives are investigated The basic requirement in determining the increase in the LCC is to fix-a risk level (for each system) at which the LCC is determined The three systems have different characteristics such as the generation mix, outage rates and the reliability criteria Table 1 illustrates typical parameters for the three systems for a given year during the system development sequence considered in the study

It can be observed from Table 1 that Manitoba system is predominantly hydro whereas: the Alberta

The Saskatchewan system has approximately 20% more thermal generation than hydro The system installed capacities, operating

0018-9510/83/0900-2826$01.00 © 1983 IEEE

reserves and F.O.R.'s are different for the three

systems The risk criteria used for generation

planning for the three systems are also different and

hence a single risk index cannot be used for all the

systems in obtaining the LCC The risk value used must

reflect the system reliability performance and

therefore, the following procedure was adopted to

select compatible risk levels for all systems

As a first step in the study each system was

examined without considering any interconnection The

risk levels corresponding to the forecast peak loads

for different years were obtained for each system based

upon the individual system generation expansion

sequence A study year was chosen during which each

system has approximately its lowest acceptable reliabi-

lity level This study year generally occurs just

prior to the addition of new generation _

The next step was to determine for each system,

for the chosen study year the risk index corresponding

to its forecast peak load considering the existing

interconnections These risk indices were then used as

the standard risk indices at which the LCC with the

grid was obtained for each system

Table I Typical System Data for the Year 1999

Manitoba Alberta |Saskatchewan Total Installed

Percentage

Hydro

Percentage

Percentage

Range of Unit

Size (MW) 2.5-118.0/3.0-360.0 10,0-280.0

Range of

F.O.R.'s 0.,005-0.03/0.001-0, 384 |0.0014~0.40

After obtaining the interconnected system risk

indices, the LCC's corresponding to these risk indices

were obtained considering the generation configuration

for each system with the grid for the chosen year The

effect of the grid was examined by comparing the LCC

values obtained with the grid and proposed transactions

to those without the grid for each system The

procedure outlined above evaluates the interconnection

benefits due to the grid for a given year and considers

the difference in reliability performance of the three

systems without the grid The modeling procedure, the

selection of risk criteria and the system studies are

described in the subsequent sections

3 SYSTEM STUDIES WITHOUT THE GRID

A schematic diagram of the system used in the study is

given in Figure 1 The proposed Western Canada Grid

interconnection which consists of a 500 kV HVDC link

between Nelson River (Northern Manitoba) generating

system to Calgary (Alberta) with an intermediate tap at

Saskatoon (Saskatchewan) is shown in dotted lines The

transmission capabilities of the ties are shown in

parentheses All the systems shown as blocks were

modeled in detail together with the transmission

systen The generators were modeled using a binary representation The transmission system consists of HVDC as well as A.C ties

It can be observed from Figure 1 that the system without the grid is a radial system with considerable inter-system connections and hence it was decided to model the system using the capacity assistance probability method The effect of the grid is to introduce a single loop The method used to consider the effect of this loop on the system reliabilities is illustrated in Section 4.2 The bulk of the Manitoba system generation in the north is transmitted via the HVDC transmission system and its outage probabilities have considerable effect on the reliabilities of the systems in the grid The HVDC transmission system was modeled as a multi-level derated unit with a number of capacity states and associated probabilities caused by valve group, pole and bipole failures’ The A.C transmission system modeling procedure is similar to the recursive method of generation addition to obtain the generation system model

NORTHERN

NORTHERN

LINK $00 KV MANITOBA

BEPC SYSTEM

100 MW TIE

| LINE 500 KV

HVỌC TA ( IO00 MW}) 8.C Cc, HYDRO

LINK 500 KV E.— ALBER

== = PROPOSED HVDC TIE-LINES wee EXISTING TIE-LINES

In these studies the standard risk indices for the three systems and a suitable study year are selected based upon which the improvement in the LCC value with the grid is obtained These studies are classified as follows:

1) Isolated System Studies 2) Interconnected System Studies

3.1 Isolated System Studies The object of these studies is to determine a study year in which each system has a risk level that approaches the design limit that each system appears

to be developing to meet

a) Manitoba System Studies

The proposed development in Manitoba consists of hydro-electric plants These plants are generally quite large relative to the annual load increases in

Manitoba and therefore there are only a few years in

which new stations are added to the system As a

result there are only a few years just prior to new

plant additions in which the risk index approaches the

maximum acceptable value The size of generation

additions in Saskatchewan and Alberta more closely

match the annual load growth, and therefore’ the

Manitoba system was selected first to find an

acceptable study year

The Manitoba Hydro system is comprised of the

Southern and Northern Manltoba systems The

intra-system connections from the Northern Manitoba to

i) A 230 kV A.C transmission line

ii) A HVDC transmission system

The Northern Manitoba load is first met by the

generators in the North not connected to the HVDC

system and the remaining capacity is available for

transmission to the South via the 230 kV line The

remaining generators in the North are available to the

Southern Manitoba system via the HVDC transmission

system The capacity outage table for the Southern

Manitoba system was obtained considering the assistance

from the Northern’ system This outage table in

conjunction with the load variation curve of the

Southern system was used in obtaining the risk index

for the Manitoba system The studies are conducted for

five development stages of generation additions The

HVDC facilities also change with the increase in

generation in the Northern Manitoba system The risk

values obtained for different years are given in

Figure 2

` l7; 9 A đ ' ø@

A Ly 7 œso-ế VN AC 2À >

z footy

a oi i! , ì

F —— SASKATCHEWAN SYSTEM £ 6 1 |

-+++ ALBERTA SYSTEM i

An MANITOBA SYSTEM Ỷ có

0.002 Jeera eee eed ell esl nelle c]Ì | |

1982 1985 1988 199} 1994 1997 2000 2002

YEAR

FIGURE 2 ISOLATED SYSTEM RISK LEVELS AT THEIR

FORECAST PEAK LOADS The risk index approaches the apparent design

limit only twice during this development period; once

in the year 2000 when it reaches 0.16 day/year and

again in 2018 when it reaches 0.23 day/year

b) Alberta System Study

The peak load vs risk index curves for the Alberta

systems were obtained using the generation and load

models for each year The risk levels for the peak

loads for different years are illustrated in Figure 2

The peak load vs risk curves for this system were obtained using the same procedure as for the Alberta system study The risk values for different years are

It can be seen from Figure 2 that the Manitoba system approaches the maximum risk index only twice The Alberta and Saskatchewan development sequences were not provided up to 2018 and therefore the preferred study year from Manitoba's viewpoint is

2000 Alberta and Saskatchewan risk indices do not approach their apparent maximum level in 2000 The year 1999 is a more acceptable year for both Alberta and Saskatchewan The Manitoba system risk index in

1999 is not much less than in 2000 and therefore the year 1999 was selected as the study year as it was the best compromise for all the three systems

3.2 Interconnected System Studies

In the previous section, the year 1999 was chosen

as the study year based on isolated system studies

In this section, a standard risk level is obtained for each system based upon the existing interconnections during the 1999 study year The basic procedure is to obtain the risk index (considering the existing inter- connections) corresponding to the forecast peak load during the 1999 study year for each system These risk levels were used in evaluating the effect of the grid on the load carrying capabilities of the individ- ual systems The interconnected system studies consist of the following three parts:

a) Manitoba Interconnected System Study b) Alberta Interconnected System Study c) Saskatchewan Interconnected System Study a) Manitoba Interconnected System Study

The Manitoba Hydro system has interconnections with the following systems:

i) Saskatchewan Power Corporation System ii) Ontario Hydro West System

iii) The MAPP System of the U.S.A

In this study all these systems were represented together with their ties The capacity outage probability table of the Southern Manitoba system was obtained considering the assistance from the Northern Manitoba system as well as the Saskatchewan, Ontario- West and the MAPP systems The peak load for the Southern Manitoba system was varied and the risk index obtained Figure 3 illustrates the risk curves for

interconnections The forecast 1999 Southern Manitoba system peak load is 4511 MW which results in an isolated system risk index of 0.11 day per year The corresponding Manitoba interconnected system risk index is 0.001 days per year This risk value was used for the Manitoba Hydro system in further studies

b) Alberta Interconnected System Study

The only interconnection that Alberta system has

is with the British Columbia Hydro system through a

1000 MW tie In this study the B.C Hydro system was represented in detail The U.S ties with the B.C Hydro system were not represented This would not affect the results as even without these U.S ties the benefits to Alberta system of the B.C tie is 1000 MW which is the maximum capacity of the tie After meeting its peak load, any assistance from the B.C Hydro system is available through the tie line, the

outage c) Saskatchewan Interconnected System Studyv Curves

The Saskatchewan system is interconnected with

limiting factors being the capacity and forced

rate of the tie line Figure 4 shows the risk

for the Alberta system with and without the inter-

the Manitoba system which in turn has interconnections connections The forecast peak load for the system

during 1999 is 11 225 MW The corresponding isolated with the MAPP and Ontario Hydro West systems In this system risk index is 0.85 day per year The risk level study all these systems were represented in detail corresponding to this load for the Alberta inter- together with their tie lines After meeting its own connected system from Figure 4 is 0.015 day per year demand, the Manitoba system provides assistance This risk level is therefore used to obtain the LCC of (through the tie lines) to the Saskatchewan system

Saskatchewan system is shown in Figure 5 The forecast peak load during 1999 is 3906 MW which results in an isolated system risk index of 31.0 days per year The risk index corresponding to this load for the Saskatchewan interconnected system is 11.5

This risk value was used as a standard

* , Ps

- o studies It can be observed from Figure 5 that the

oIR— z ? / : | approximately 200 MW indicating that the additional

, : ` ° «

, /

— ; / —

# , ể,

0000! lộ Ì FÀ ’ | | I | 10.0 >> z #

~ = , ia

La , if

eof

— z ra

/ a Š - ; i}

“ ‘2 mw , i;

— z “es , if

, , 2 | ; if

“ af 0.0l

# ff

~— | / tự :

” r #ể

: FF ff snneeneee INTERCONNECTED SYSTEM systems corresponding to the forecast peak loads

(0.000 II 000 12000 13000 System (days/year) (Forecast Peak Load in 1999) _ ALBERTA LOAD (MW)

FIGURE 4 ALBERTA SYSTEM RISK CURVES |

Saskatchewan

4 WESTERN CANADA GRID STUDIES

In this section the effect of the proposed Western

Canada Grid on the reliabilities and LCC's of the three

Systems is investigated for the year 1999 The

schematic diagram of the system used in these studies

is shown in Figure 1 The proposed Western Canada Grid

interconnection consists of a 500 kV HVDC transmission

system from the Northern Manitoba system to the Alberta

system with an intermediate tap at Saskatoon linking

the Saskatchewan system This interconnection provides

a 500 MW sale to the Saskatchewan system and a 1000 MW

sale to the Alberta system from the Manitoba Hydro

system The hydroelectric generation in Northern

Manitoba was increased to provide for this total sale

of 1500 MW The generating capacities of Saskatchewan

and Alberta systems were reduced by the amount of

purchase from the Manitoba Hydro system A schematic

diagram of the proposed ties is given in Figure 6 The

grid introduces a loop containing the Northern

Manitoba, Southern Manitoba and Saskatchewan systems

In order to analyze the systems with the grid, the

effect of this loop on the system reliabilities and a

method of considering it must first be investigated

There is no direct tie between the Southern Manitoba

and Alberta systems and hence as a first approximation

the effect of the Southern Manitoba system can be

neglected when studying the Alberta system It has

been found in Section 3.2 that the effect of

considering the Southern Manitoba system on the

Saskatchewan system LCC is to increase the value by

approximately 200 MW The approximate LCC values of

the Saskatchewan system can be obtained by increasing

the LCC values obtained by ignoring the Southern

Manitoba system by 200 MW The approximate upper bound

of Manitoba system LCC can be obtained by assuming a

100% reliable generation in the Northern Manitoba

system and considering the HVDC tie between’ the

Northern and Southern Manitoba systems The LCC's

obtained using these approximate models were compared

with those using more detailed representations

NORTHERN MANITOBA

HVDC LINK 4x450 MW VALVE GROUPS

IOOOkm | 500 kV LINE

VALVE GROUPS BUT TOTAL CAPACITY ASSUMED LIMITED TO 500 MW

566 km HVDC LINK 2x500 MW VALVE GROUPS TOTAL CAPACITY = 1000 MW

ALBERTA

FIGURE6 PROPOSED HVDC TIES FOR

THE WESTERN CANADA GRID 4.1 System Studies Using Approximate Models

The studies illustrated in this section were based

on approximate models The purpose of _ the

(1) In order to consider the loop formed by the Northern Manitoba, Southern Manitoba and Saskatchewan systems using the capacity assistance probability method, decoupling between the systems must _ be established The approximate models helped in identifying for example, the effect of neglecting the Southern Manitoba system on the LCC values of Alberta and Saskatchewan systems

(2) To estimate the values of loads to be used in the other two systems while analyzing each system As an example the LCC value of Saskatchewan system first estimated using the approximate model was used in the subsequent studies of the Alberta and Manitoba systems This procedure ensures that the risk index

of the Saskatchewan system is maintained the same with and without the grid when obtaining the LCC's of the other systems

A number of cases were studied in order to assess the impact of the grid on the LCC's of the three systems Approximate results for the LCC values of the Saskatchewan and Alberta systems were obtained by allocating 10 units of Limestone generating station (each of 108 MW) and 6 units of Conawapa station (each 127.4 MW) for a total installed Northern Manitoba system capacity of 1844.4 MW to meet the 1500 MW sale

by the Manitoba system and ignoring the rest of the Manitoba system The LCC's of Alberta and Saskatchewan systems were obtained first without considering the Southern Manitoba assistance The LCC's considering the Southern Manitoba system were then obtained using the capacity assistance probability table of the Southern Manitoba system The grid tap at Saskatoon on the HVDC system consists

of 4 valve groups each of 250 MW capacity for a total capacity of 1000 MW The total available capacity, however, was limited to 500 MW due to design limitations This capacity can be increased to 1000

MW by providing additional equipment at an increased system cost The effect of this increase in the tap size to 1000 MW without increasing the power export to the Saskatchewan system was also investigated

The LCC values obtained using the approximate models are compared with those using more detailed models in Table II presented in the next section 4,2 System Studies Using Detailed Models

A more detailed representation was used to verify the results obtained using the approximate models The first system used in the studies was the Southern Manitoba system Later this model of the Southern Manitoba system was used in obtaining the LCC's of the Saskatchewan and Alberta systems

(a) Manitoba System Study

the Manitoba following The load

system was

carrying capability of

- procedure:

The Northern Manitoba system was represented as

in the proposed grid system and the total sale of 1500

MW was represented as a firm load The assistance from Alberta and Saskatchewan systems were represented

by their capacity assistance probability tables The HVDC transmission system between the Northern and Southern Manitoba systems as well as the Southern Manitoba system and its ties with the MAPP and Ontario systems were represented in detail The Ontario West and MAPP systems were represented in detail and the

LCG value of the (Southern) Manitoba system at a risk

index of 0.001 day per year was obtained as 4841 MW This value is 2 MW less than that obtained assuming 100% reliable generation in the North in the previous section It can be concluded that the upper bound LCC

was realized with the generation in the Northern

Manitoba as proposed for the grid and any increase in

the Northern generation will not significantly improve

the LCC of the Manitoba system This is mainly due to

the fact that the HVDC link between the Northern and

Southern Manitoba systems limits the possible

assistance from the Northern system to the South to

3668 MW which is the maximum capacity of the tie The

load carrying capability of the Manitoba system

obtained using the detailed representation is very

close to the value obtained using the approximate model

of the previous section and this value is the upper

bound value for the year 1999 The Manitoba system

risk curve with the grid is illustrated in Figure 3

b) Saskatchewan System Study

In order to check the LCC

Saskatchewan and Alberta systems obtained in the

previous section, the assistance from the Southern

Manitoba system to the North was obtained as follows:

values for the

The Southern Manitoba system and its inter-

connections with the grid for the study year were

represented as in the Manitoba system study described

previously The load in the Southern Manitoba system

was assumed to be the LCC value obtained for that

system with the grid (4841 MW) The assistance from

the Southern Manitoba system to the Saskatchewan system

was represented as a firm load of 200 MW on the

Southern Manitoba system The potential assistance

from the Northern Manitoba to the South was assumed to

be 3668 MW with 100% reliability The Southern

Manitoba system assistance after meeting a load of

(4841 + 200) = 5041 MW was passed through a firm tie of

3668 MW and the resultant capacity assistance

probability table developed The assistance table

developed using this procedure maintains the risk level

at the same value as without the grid and allows the

Manitoba system load to be at its maximum value This

procedure yields conservative LCC values The same

capacity assistance table was used in studying

Saskatchewan and Alberta systems

In this study, the Northern Manitoba system was

represented in detail A firm load of 3668 MW was

placed in the Northern Manitoba system to represent the

assistance to the Southern system The assistance from

the Southern system back to the North, represented by

the capacity assistance probability table obtained

previously, was injected into the Northern Mantoba

system This representation ensures that the assist-

ance that can be given to the North is the capacity

left after meeting the Southern system load A firm

200 MW was injected into the Saskatchewan system to

represent the assistance from the Southern Manitoba

system The Alberta system assistance probability table

was the same as that used in the previous section The

capacity of the Saskatoon tap was varied from 500 to

1000 MW and the LCC values were calculated The LCC

values were 3936 MW and 4413 MW for Saskatoon tap

capacities of 500 and 1000 MW's respectively The

Saskatchewan risk curve with the grid is included in

.Figure 5

c) Alberta System Study

_ The procedure used for obtaining the Alberta

system LCC with the grid is similar to the one used for

the Saskatchewan system study except that the Alberta

system replaces the Saskatchewan system The

Saskatchewan system assistance probability table used

in this study has the Southern Manitoba -system

assistance of 200 MW injected as 100% reliable

generation This was done to account for the 200 MW

assistance provided by the Southern Manitoba system to

Saskatchewan system The LCC value of the Alberta

system obtained from this study at a standard risk level of 0.015 day/year is illustrated in Figure 4

To summarize, the LCC values for the three systems in the proposed Western Canada Grid for the

1999 study year are shown in Table II

Table II Comparison of the LCC Values with the Grid

Using Approximate and Detailed Models

Load Carrying Capability (MW) System Detailed Estimate Approximate Estimate

Saskatchewan

Table III illustrates the effect of the grid on the LCC values and risk indices of the three systems The installed capacities with and without the grid were obtained from the generation development sequences | provided by the utilities In the study year, with the grid, Manitoba Hydro adds 1466 MW additional

‘generation to provide for the total sale of 1500 MW to Alberta and Saskatchewan Alberta and Saskatchewan systems reduce their generation by 1125 and 532 MW respectively while still maintaining approximately the same risk index as without the grid The Manitoba

‘risk index with the grid and the 1466 MW additional generation to supply the 1500 MW export is improved significantly from its risk index without the grid Thus there is a net overall reduction in generation of

191 MW with the grid The net increase in LCC for the three systems, if their respective risk indices were maintained at exactly the same indices as without the grid, is 291 MW The overall benefit with the grid due to decrease in installed capacity: and increase in LCC is therefore 482 MW

4.3 Effect of Saskatoon Tap Capacity on the System

‘the Saskatchewan

Reliabilities

The the HVDC LcC's of

studies described previously indicated that transmission system in the grid limits the the Saskatchewan and Manitoba systems The HVDC tie between the Northern and Southern Manitoba systems constrains the LCC of the Manitoba system whereas the grid tap at Saskatoon limits the LCC of

system The capacity of the Saskatoon tap has considerable effect on the reliability performance of the Saskatchewan system and because of the interconnection has an effect on the Manitoba system The effect of increasing the tap capacity on the reliabilities of these systems was investigated

a) Saskatchewan System Reliability

The Saskatoon tap capacity in the proposed grid interconnection was assumed to be 500 MW This is exactly equal to the amount of power export from Manitoba system to the Saskatchewan system As noted previously, the capacity of this tap can be increased

to 1000 MW This increase in capacity would provide enough capability for the 500 MW import from the Northern Manitoba system plus any assistance that is available from the Alberta system The risk index corresponding to the 1999 peak load decreases to 0.54 day/year with 1000 MW tap from 9.2 days/year with a

500 MW tap Thus there is a marked improvement in

reliability with the increased tap capacity to the

Saskatchewan system

Table III - Reliability Study of Western Canada Grid With 500 MW Saskatoon Tap

1, Risk Index for the individual system forecast load

Capacity Peak Load | Capacity L.C.C.2] Without With Capacity jin L.C.C Total

2 LCC with the grid at the risk level which would exist without the grid

The effect of increasing the Saskatoon tap

capacity on the Manitoba system reliability was

investigated by repeating the Manitoba system study

with a 1000 MW Saskatoon tap The risk index obtained

in this case was 0.0013 day/year The corresponding

value with a 500 MW tap was 0.0003 day/year The

reduction in the risk indéx in this case is not as

significant as for the Saskatchewan system

The reliability benefits can be assessed in two

alternative ways as follows:

If the risk index is maintained the same without

and with the grid, then there is an improvement in the

Saskatchewan system with a 500 MW tap is 3936 MW The

corresponding LCC value with a 1000 tap is 4434 MW

This is an increase of 498 MW for this system and

almost all the increase in the tap capacity (500 MW) is

available as an increase in the LCC Similar results

for Manitoba system showed that the increase in LCC

with 1000 MW tap was 250 MW

Alternatively, if the system peak load 18

maintained at the same value with and without the grid,

the risk level decreases with the grid The effect of

increased tap size on the risk indices for the

Saskatchewan and Manitoba systems was illustrated in

the beginning of this section which shows that there is

a significant reduction in the risk index for the

Saskatchewan system and a less significant reduction

The above mentioned interconnection benefits of

improved LCC and reduced risk index cannot both be

obtained simultaneously without adding new generation

5 SUMMARY AND CONCLUSIONS

reliability

Canada Grid

This paper has illustrated’ the

benefits associated with the Western

consisting of a direct tie

generating system to Calgary with an intermediate tap

at Saskatoon These studies in addition to the usual

modeling of the generating units considered the

interconnections with the neighboring systems and the

constraining effects of the HVDC ties for all the

systems forming the grid The study addressed only the

capacity and the resulting steady-state adequacy of the

systems and did not consider the energy requirements or

the economic alternatives The study also did not

consider the optimum location for reserve in the three

systems Based on the study reported in this paper,

the following conclusions can be made

- The proposed Western Canada Grid

steady-state generation system adequacy for the

three systems and hence they could reduce the

amount of reserve ‘capacity from that - planned

from the Nelson River

improves’ the

without the grid and still maintain the

- The HVDC transmission system constraint and reduced the achievable interconnection reliability benefits To fully take advantage of the ties, the capability of the HVDC transmission system out of Northern Manitoba must be increased from that assumed in the study

- Increasing the capacity of the tap at Saskatoon provides an opportunity of both relieving the transmission constraints out of Northern Manitoba and of improving the reliabilities of the Saskatchewan and Manitoba systems

These conclusions are based upon limited studies

of the system for one typical year, but indicate that

in future reliability studies of the grid the major emphasis must be placed on the capacity, reliability and the routing of the HVDC transmission facilities out of Northern Manitoba The effect of the grid on the system reliabilities for the other years and the sensitivity of the risk indices to variations in the HVDC transmission system parameters such as the configuration, capacity and outage data of the components require further investigation

same

acted as oa

6 REFERENCES

1, R Billinton, Power System Reliability Evaluation (Book) Gordon and Breach Science Publishers: New York, 1970

2 M.P Bhavaraju, "Application of - Probability Techniques in the Evaluation of Generating Capacity Reliability in Single and in Two-Interconnected Systems", M.Sc Thesis, College of Graduate Studies, University of Saskatchewan, 1967

3 R Billinton, M.P Bhavaraju and P.R Thompson,

"Power System Interconnection Benefits", Transactions of the Canadian Electrical Association Vol 8, Part III: 1969,

4 C.K Pang and A.J Wood, "Multi-Area Generation System Reliability Calculations", LEEE Transactions on Power Apparatus and Systems,

Vol PAS-94, No 2, pp 508-517: March/April

5 W.A Derry, P.R Thompson and E.A Zaleski,

"Quantifying Economic Benefits of Improved System Reliability Due to Interconnections" Transactions

of the Canadian Electrical Association Engineering and Operating Division Vol 8l, Part IIL, Paper No 79-SP-162: 1979

Discussion Stephen T Lee and Chok K Pang (Energy Management Associates,

Inc., Santa Clara, CA): Interest in reliability indices has been increasing

among utilities in the U.S Unfortunately, not much publication is

available on the actual numerical reliability indices of the U.S utilities

The authors have contributed to the data base of typical utility reliabili-

ty indices with this paper

In our consulting activities, we often have to deal with the difficult

question of ‘‘What is the appropriate level of reliability for generation

and interconnection planning?’’ Using a base case resource plan as a

reference, and then evaluating the load carrying capability of an alter-

native resource plan or an interconnection scheme to achieve the same

relaibility level as in the reference case is not an entirely satisfactory ap-

proach A utility planner wil still need to justify to the regulatory agen-

cies and to the customers why that utility uses a LOLP criterion of x

days per year Would the authors comment on the numerical LOLP

criteria that are being used or are being accepted by the Canadian

utilities? Would the authors also suggest ways of establishing or defen-_

ding a reliability criterion for a particular utility?

Manuscript received February 23, 1983

P.R S Kuruganty, P R Thompson, and R Billinton: We would like

to thank the discussers for their comments We hope that we have made

a contribution to the data base of actual utility applications in quan-

titative reliability evaluation The question that Lee and Pang raise is

regarding the optimum reliability level to satisfy the customers’ demand

as well as the economic constraints This depends upon many aspects

such as the system installed capacity, generation mix, type of intercon-

nects and agreements with neighboring systems regarding emergency

assistance as well as the system load charcteristics The basic problem in

system planning is to develop a system which will satisfy the customers’

load requirements at a reasonable level of reliability and at an accep-

table cost The system planner must consider the customer demand for

quality of service and the economic limitations of providing a highly

reliable system Recently there has been an increasing interest in incor-

porating customer outage costs in power system reliability planning A

number of papers have been published on this subject Many North

American utilities presently use of loss of load expectation (LOLE) in-

dex of 1 day in 10 years This translate into a loss probability (LOLP) of

0.3 x 10-3 to 0.45 x 10-3 depending upon the assumptions regarding

the load model and the method of calculation It should be realized that

there are many methods (with attendent assumptions) to calculate the

LOLE index and therefore the LOLE of x days/year does not necessari-

ly reflect the same level of reliability in all utilities using the same numerical value of LOLE In choosing any risk index for practical systems on has to use engineering judgement Most applications assume that all the generation is at one bus to which the system load is con- nected This, in effect, means that the outages of transmission network were completely ignored This is not realistic in systems with a signifi- cant intra-system transmission network We have briefly described the rationale behind the LOLE criterion used at Manitoba Hydro The Manitoba Hydro system has a considerable intra- and inter-system transmission network The bulk of our generation in the Northern system is transmitted to the South by a HVDC link The effect of this link on the overall system reliability cannot be ignored In addition, the inter-system transmission between Manitoba Hydro and the neighbor- ing systems must be considered We model all these systems in our LOLE computations Past performance records indicate that a 12% reserve of hydro generation was quite adequate for system planning | This translates into a LOLE of 0.1 day/year for Manitoba Hydro system in isolation (without considering inter-system ties but consider- ing intra-system transmission) When the interconnections are con- sidered this risk index decreases to 0.003 day/year We have, therefore, two risk indices that are considered in system planning, viz 0.1 day/year without considering external ties and 0.003 day/year with in- terconnections taken into account Regarding the reliability criteria us-

ed by other Canadian Utilities, these are described in detail in Reference

In conclusion the philosophy of choosing the desirable level of reliability at Manitoba Hydro is justified by the following argument In choosing the required level of reliabiliy one uses judgement We have examined the past operating record and the reliability levels of the system We are satisfied with the past record of reliability of supply and the associated reserve levels We continue to monitor the effects of system changes on the reliability and associated cost with a view to modify the reserve and/or intra-system transmission requirements if considered to be warranted We thank Drs Lee and Pang for their discussion

REFERENCES [1.] R Billinton, ‘Reliability Criteria Used by Canadian Utilities in Generating Capacity Planning and Operation’, JEEE Transac- tions PAS-97, 1978, pp 984-989

Manuscript received April 18, 1983