Mixed GA-OPF based optimal procurement of energy and operating reserve in deregulated environment

Abstract

Ancillary Services (AS) plays a vital role in a deregulated environment because these services act as the frontier of a power system. It helps to maintain the quality and safety of the supply. Operating Reserve (OR), as an important AS, has been considered in this work. This paper proposed a mixed Genetic Algorithm (GA)-Optimal Power Flow (OPF) mechanism can act as an effective tool for procurement of different services like energy and AS. The sequential clearing technique has been considered for procurement of Energy and OR an objective of cost minimization. Herein, the EM is cleared first in Energy Market (EM) followed by clearing of Operating Reserve Market (ORM). The proposed approach for obtaining the required service using mixed GA-OPF approach has been investigated by considering modified IEEE-30 bus test system.

Keywords

Electric power deregulation genetic algorithm optimal power flow operating reserve disaggregated approach energy and AS market

Abbreviations

Available Capacity

Ancillary Services

ARC

Available Reserve Capacity

Energy Market

Forward Markets

Genetic Algorithm

IDM

Intra-Day Markets

ISO

Independent System Operator

JOD

Joint Optimization Dispatch

FERC

Federal Energy Regulatory Commission

MOD

Merit Order dispatch

OPF

Optimal Power Flow

ORM

Operating Reserve Market

RTM

Real-Time Market

SQD

Sequential Dispatch

TMNSR

Ten Minute Non Spinning Reserve

TMOR

Thirty Minute Operating Reserve

TMSR

Ten Minute Spinning Reserve

1 Introduction

To enhance competition, power system utilities have participated in the process of restructuring. This reformation of the power sector brought significant changes in power system operation [17]. Figure 1 shows the disaggregation of the vertically integrated utilities into several independent entities like Generating Companies into GENCOs, Distribution Companies into DISCOs, and transmission Companies into TRANSCOs.

Fig.1

Disaggregation of traditional vertically integrated utility.

Reformation in the electrical sector, primarily focuses on increasing system efficiency with improvement in service standards and cost minimization by developing a competitive marketplace [23].

2 Market mechanism for energy and ancillary services

The market mechanisms for the procurement of energy and associated services in deregulated market has been discussed in this section. These mechanisms reveal different possibilities of finding the accuracy among the different results obtained for viability and functioning of efficient markets [8].

2.1 Market structure design

Forward and Real-Time Markets (RTM) are the important market structure designs all around the world for energy and/or AS dispatch and is shown in Fig. 2 [1].

Fig.2

Energy and ancillary services dispatch auction market design.

Forward Markets (FM) operate at definite time before the actual delivery of the services [15]. This may either be 24 hours-ahead or hour-ahead timeline. As a component of FM, Intra-Day markets are used for making adjustments obtained from the previous settlements by allowing market participants to re-bid into the market before the actual delivery [13].

The RTM is used to maintain generation-demand equilibrium during real-time operation. The requirement of these markets is to obtain the requisite level of the reliability. For this, the system operator must acquire desired services [2]. Maximum electricity markets of developed markets (the US, Australian National Electricity Market, etc.) as well as markets of developing nations have such type of market.

2.2 Market clearing mechanism

The basic form of clearing technique is Merit Order Dispatch (MOD), in which the market is cleared on the basis of lower expensive offer for energy and AS [9]. None can guarantee the practicability of the result when the commodities are coupled [19].

Sequential dispatch, as an expansion of MOD identifies the truth that several services consume the same resource. This technique clears the markets on the basis of priority sequence defined for each commodity [3, 7]. In Joint Optimization Dispatch (JOD), different commodities are cleared simultaneously in a same market [16]. The justification of the schedule and pricing in this technique is hard, but clearly model the dominant coupling between the commodities [10].

Evolution of auction mechanism for energy and AS markets worldwide is shown in Fig. 3.

Fig.3

Evolution of auction mechanism for energy and AS markets.

3 General description of ancillary services

The NERC, a non-profit international regulatory authority utilizes AS to assure the reliability of transmitting bulk power from resources to loads in North America [18]. In England and Wales, National Grid Company ISO utilize AS for two roles, balances supply-demand; and limit frequency and voltage to remain within the ranges [12]. In the USA, FERC Directive 888 defines 12 technical and non-technical AS.

The simplest classification of the resource based AS helps us to group them into frequency control services (like regulation, load following, operating reserves), voltage control services (through reactive power support) and emergency services (by black-start services) [20], as shown in Fig. 4.

Fig.4

Resource based ancillary services.

The characteristics of OR is discussed in the present section. OR is planned to respond to uncertainties and are needed for sustaining the security of the electrical scheme [4]. It primarily deals with unscheduled generation outages and disruptions. Depending upon the deployment time, OR can be further divided into-

Ten Minute Spinning Reserve (TMSR): It is supplied from the resources that are grid-synchronized, partially loaded, and is entirely responsive within ten minutes of an ISO instruction.

Ten Minute Non Spinning Reserve (TMNSR): Similar to TMSR, but need not to be grid-synchronized, must be accomplished within ten minutes and capable of sustaining the response for at least 120 minutes [9].

Thirty Minute Operating Reserve (TMOR): This reserve must be available completely in 30 to 60 minutes, depending on the area, and must be capable of sustaining that response for 120 to 240 minutes. On-line and off-line generation, and interrupted load can provide this reserve [11].

Herein, AS procurement relates to the procurement of the TMSR, which comes under OR, as described above.

4 Proposed mixed GA-OPF based optimization framework for EM and ORM dispatch

The work considers that the auctions for energy and OR are operated by the ISO to conform the system reliability obligations. Minimizing the revenues paid to the participants for supplying the required service (energy and/or AS) is the main objective of the dispatch.

This work considers linear bidding coefficients from the participants in terms of amount and the corresponding prices in EM and ORM. Generator-side bidding is considered in the present work. Each supplier submits individual bidding blocks in EM and ORM as [14.]:

In EM - $GENCOs : E_{si}^{j}, {PE}_{si}^{j}; j = 1, 2, 3, \dots . ., n_{esi}$ (1)

In ORM - $GENCOs : R_{si}^{l}, {PR}_{si}^{l}; l = 1, 2, 3, \dots . ., n_{rsi}$ (2)where $E_{si}^{j}$ and ${PE}_{si}^{j}$ is the quantity and price of energy offered by i^th supplier in j^th band in EM whereas $R_{si}^{l}$ and ${PR}_{si}^{l}$ is the quantity and price of reserve capacity offered by i^th supplier in l^th in ORM.

4.1 Payment mechanism

The payment mechanism in the energy and OR markets is discussed under this section. The payment mechanism is considered to be Pay-As-Bid (PAB) [5, 24] in both EM and ORM as:

4.1.1 In Energy Market (EM)

$\begin{array}{l} E P_{s i} (E_{s i}) & = \sum_{j = 1}^{a} E_{s i}^{j} (P E_{s i}^{j}) \\ + (E_{s i} - \sum_{j = 1}^{a} E_{s i}^{j}) P E_{s i}^{a + 1} \\ \sum_{j = 1}^{a} E_{s i}^{j} \leq E_{s i} \leq \sum_{j = 1}^{a + 1} E_{s i}^{j} \end{array}$ (3)

4.1.2 In OR Market (ORM)

The payment to the accepted reserve (R_si) will be- $\begin{array}{l} R P_{s i} (R_{s i}) & = \sum_{l = 1}^{c} R_{s i}^{l} (P R_{s i}^{l}) \\ + (R_{s i} - \sum_{l = 1}^{c} R_{s i}^{l}) P R_{s i}^{c + 1} \\ \sum_{l = 1}^{c} R_{s i}^{l} \leq R_{s i} \leq \sum_{l = 1}^{c + 1} R_{s i}^{l} \end{array}$ (4)

4.2 The proposed algorithm and the objective function

To minimize OR payments to suppliers, sequential approach for EM and ORM has been implemented. In this flexible mode of operation both energy and reserve are optimized respectively under GA and MATPOWER environment by dispatching the complete capacity of the unit. Therefore, the mechanism is illustrated through a step-by-step algorithm:

Step 1. By using GA, clear EM from available suppliers by

Collect the bus data, line data and cost coefficients and their limits,

Convert and optimize the constrained optimization problem using GA technique [25].

$Min \sum_{i = 1}^{N_{s}} {PCE}_{si} (E_{si})$ (5)

such that $\sum_{i = 1}^{N_{s}} E_{si} = E_{L} + E_{Losses}$ (6)

Step 2. After clearing of EM from available suppliers using GA, TMSR is dispatched from the ARC of the individual unit in ORM. AC for a i^th supplier in ORM is calculated from Equation (7)[6, 22]. ${AC}_{i} = E_{si}^{Max} - E_{si} i = 1, 2, \dots . N_{s}$ (7)

AC and ARC in ORM are calculated by using following relations: ${({ARC}_{i})}_{TMSR} = Min [{(AC)}_{i}, τ \times {RR}_{i}]$ (8)where i = 1, 2, … N_s,

N_s: Number of resources available in ORM

RR_iramp rate of i^th supplier

τ: Specific response time for ORM

(hereforTMSR, τ = 10)

Step 3. Similar to EM, clear the ORM under MATPOWER environment such that the procurement cost of obtaining TMSR from available suppliers is minimized [21]. Mathematically $Min \sum_{i = 1}^{N_{s}} {PCR}_{si} (R_{si})$ (9)

such that $\sum_{i = 1}^{N_{s}} R_{si} = R_{L} + R_{Losses}$ (10)

The minimization of procurement costs in Equations (5) and (9) is subject to the generation and transmission constraints in Equations (11–15) [5]:

The power flow equation of the power network $f (V, φ) = 0$ (11)

The inequality constraint on real power generation P_gi at PV buses $P_{gi}^{\min} \leq P_{gi} \leq P_{gi}^{\max}$ (12)

The inequality constraint on reactive power generation Q_gi at PV bus I $Q_{gi}^{\min} \leq Q_{gi} \leq Q_{gi}^{\max}$ (13)

The inequality constraint on voltage magnitude V_i at each PQ bus $V_{i}^{\min} \leq V_{i} \leq V_{i}^{\max}$ (14)

MVA flow limit on transmission line ${MVAf}_{ij} \leq {MVAf}_{ij}^{\max}$ (15)

5 System characteristics for validation of proposed approach

To authenticate the efficacy of the work, an IEEE-30 bus test system is modified whose details are discussed in this section.

5.1 Characteristics of generation system

Figure 5 shows the one-line diagram of the modified test system. It consists of 6 suppliers located, respectively, at buses 1, 2, 13, 22, 23 and 27. The capacities of different resources participating in the market range between 20 and 300 MW.

Fig.5

One line diagram of modified IEEE-30 bus test system.

For procuring energy from EM, the modified system has five loads E₁, E₂, E₃, E₄, and E₅ located respectively at buses 1, 2, 3, 10, and 23. These energy loads E₁, E₂, E₃, E₄, and E₅ shares respectively 10, 25, 25, 20, and 20 of the complete energy demand such that $E_{L} = E_{1} + E_{2} + E_{3} + E_{4} + E_{5}$ (16)

Technical characteristics of the suppliers, includes individual bidding blocks for EM and ORM, maximum capacities and associated Ramp Rates (RR), and are listed in Table 1.

Table 1

Technical Characteristics of Generation System

Bus	Supplier	Energy Offer ($/MWh)	Reserve Offer ($/MWh)	RR	Maximum Capacity
1	S1	40	40	5	260
2	S2	35	42	6	300
13	S3	45	54	3	20
22	S4	68	84	3	100
23	S5	56	67	4	170
27	S6	86	103	8	150

5.2 Energy and TMSR profile

During the day and in the early evening, the total energy demand will generally be higher. Similarly, the scenario of energy demand is less in the late evening and during the early sunrise as shown in Fig. 6 and is adopted from [4]. The modified system has a single load for TMSR requirement, that is, R_L located at bus 10.

Fig.6

Hourly energy demand in EM and ORM.

6 Simulation results and discussion

In the proposed approach, EM is cleared from the full capacity of the unit. After EM clearing, ORM is cleared according to Equation (8). To meet both the energy and OR, the suppliers’ schedules over the timeframe of one complete day is shown in Figs. 7–12.

Fig.7

Mixed GA-OPF based Supplier S1 schedule in EM & ORM.

Fig.8

Mixed GA-OPF based Supplier S2 schedule in EM & ORM.

Fig.9

Mixed GA-OPF based Supplier S3 schedule in EM & ORM.

Fig.10

Mixed GA-OPF based Supplier S4 schedule in EM & ORM.

Fig.11

Mixed GA-OPF based Supplier S5 schedule in EM & ORM.

Fig.12

Mixed GA-OPF based Supplier S6 schedule in EM & ORM.

It is seen from the Figs. 7–12, that economical suppliers were being already selected in EM for supplying energy to the load and are no longer available in ORM for dispatching TMSR. This is because the units with a less expensive offers, are easily accepted during the dispatching process in the EM. Since suppliers S1 and S2 are least costlier units being easily selected in the EM and will not available to deliver its capacity in ORM in some hours. During these periods, TMSR will now be dispatched using the remaining capacities at each supplier in ORM after scheduling their capacity in EM. These remaining capacities are generally available on expensive units resulting in higher procurement cost of reserves.

7 Conclusion

Competitive electricity markets require competitive procurement of energy and AS. This paper has presented a mixed GA-OPF based mechanism for procuring both energy and OR under a competitive environment by considering a sequential optimization approach. The design of such markets for energy and OR presents an objective of procurement cost minimization. Simplicity and transparency of clearing markets are the striking features of this approach. Hence the sequential optimization approach with proposed mixed GA-OPF technique provides an efficient solution while reducing the total procurement cost.

References

Banshwar

, Sharma

N.K.

, Sood

Y.R.

and Shrivastava

, Optimal location and rating of wind power plants in competitive electricity market, Journal of Renewable and Sustainable Energy9 (2017), 043306.

Banshwar

, Sharma

N.K.

, Sood

Y.R.

and Shrivastava

, Renewable Energy Sources as a New Participant in Ancillary Service Markets, Energy Strategy Reviews18 (2017), 106–120.

Banshwar

, Sharma

N.K.

, Sood

Y.R.

and Shrivastava

, A Sequential Optimization Approach for Competitive Procurement of Energy and Ancillary Services, IEEE International Conference on Electrical Power and Energy Systems (2016), 498–503.

Banshwar

, Sharma

N.K.

, Sood

Y.R.

and Shrivastava

, Market Based Procurement of Energy and Ancillary Services from Renewable Energy Sources in Deregulated Environment, Renewable Energy101 (2017), 1390–1400.

Banshwar

, Sharma

N.K.

, Sood

Y.R.

and Shrivastava

, Real time procurement of energy and operating reserve from Renewable Energy Sources in deregulated environment considering imbalance penalties, Renewable Energy113 (2017), 855–866.

Costa

A.L.

and Costa

A.S.

, Energy and ancillary service dispatch through dynamic optimal power flow, Electric Power Systems Research77 (2007), 1047–1055.

Zheng

and Zhou

, A design for regional ancillary services auction markets in China, IEEE PowerTech Conference (2003), 704–708.

Davatgaran

, Saniei

and Mortazavi

, Optimal bidding strategy for an energy hub in energy market, Energy148 (2018), 482–493.

Hirst

, Maximizing generator profits across energy and ancillary-services markets, The Electricity Journal13 (2000), 61–69.

10.

Galiana

, Bouffard

, Arroyo

and Restrepo

, Scheduling and Pricing of Coupled Energy and Primary, Secondary, and Tertiary Reserves, Proceedings of the IEEE93 (2005), 1970–1983.

11.

Sioshansi

, Evolution of Global Electricity Markets, Elsevier Science & Technology, Waltham, MA, 2013.

12.

Strbac

, Trading electricity and ancillary services in the reformed England and wales electricity market, IEEE Power Engineering Society Winter Meeting (2001), 41–44.

13.

Singh

, Auctions for Ancillary Services, Decision support systems24 (1999), 183–191.

14.

Singh

and Papalexopoulos

, Competitive procurement of ancillary services by an independent system operator, IEEE Transactions on Power Systems14 (1999), 498–504.

15.

Wang

, Shahidehpour

and Li

, Contingency-Constrained Reserve Requirements in Joint Energy and Ancillary Services Auction, IEEE Transactions on Power Systems24 (2009), 1457–1468.

16.

Afshar

, Ehsan

, Fotuhi-Firuzabad

and Amjady

, A method for reserve clearing in disaggregated model considering lost opportunity cost, Electrical Power System Research78 (2008), 527–538.

17.

Bhattacharya

, Bollen

and Daalder

, Operation of restructured power systems. Boston: Kluwer Academic Publishers, 2001.

18.

Long-Term Reliability Assessment, 2017. https://www.nerc.com/pa/RAPA/ra/Reliability%20Assessments%20DL/NERC_LTRA_Final.pdf.

19.

González

, Villar

, Díaz

and Campos

, Joint energy and reserve markets: Current implementations and modeling trends, Electric Power Systems Research109 (2014), 101–111.

20.

Raineri

, Rios

and Schiele

, Technical and economic aspects of ancillary services markets in the electric power industry: An international comparison, Energy Policy34 (2006), 1540–1555.

21.

Zimmerman

, Murillo-Sanchez

and Thomas

, MATPOWER: Steady-State Operations, Planning, and Analysis Tools for Power Systems Research and Education, IEEE Transactions on Power Systems26 (2011), 12–19.

22.

, Sun

and Cheung

, Energy and reserve dispatch in a multi-zone electricity market, IEEE Transactions on Power Systems14 (1999), 913–919.

23.

Sood

Y.R.

, Padhy

N.P.

and Gupta

H.O.

, Wheeling of power under deregulated environment of power system-a bibliographical survey, IEEE Transactions on Power Systems17 (2002), 870–878.

24.

Ren

and Galiana

, Pay-as-Bid versus Marginal Pricing–Part II: Market Behor Under Strategic Generator Offers, IEEE Transactions on Power Systems19 (2004), 1777–1783.

25.

, Dong

and Wong

, Genetic algorithm-based spinning reserve dispatching in a competitive electricity market, International Journal of Systems Science35 (2004), 847–853.