Congestion Management in Deregulated Power System using Price based Programs

Transcription

1 ol. 5 Issue 08, August-2016 Congeston Management n Deregulated Power System usng Prce based Programs. Mounka, Dept. of Eee Gvpcoe (A), sakhapatnam. Abstract--In deregulated market envronment, congeston management plays an mportant role n power system operaton. An approach of applyng Demand Response (DR) programs has been used for transmsson lne congeston management n a deregulated power system. In ths paper DR s modelled consderng Tme of Use (TOU), Crtcal Peak Prcng (CPP) usng MATPOWER software. The paper evaluates DR effects on the generatng companes, consumers, merchandsng surplus, power system securty and operatng cost n addton to the congeston management. The proposed models are mplemented on IEEE-14 bus system. Keywords - Congeston Management, Merchandsng Surplus; Demand Response, TOU, CPP, MATPOWER. I. INTRODUCTION The electrc supply ndustry has been changed from vertcally ntegrated to restructured power system. The electrcty cannot be stored n bulk easly and the transportaton of electrcty s constraned by physcal laws whch have to be satsfed to mantan the relablty and securty of the power system. In restructured market envronment, every buyer wants to buy power from the low cost generator avalable. The transmsson system has a lmted capablty to transfer power whch may overload certan transmsson lnes. Congeston referred to as a transmsson lne httng ts maxmum lmt. Transmsson congeston occurs when the transmsson capacty s nsuffcent to accommodate all the transactons. Congeston may occur due to the lack of coordnaton between generaton and transmsson companes n present scenaro. In peak perods, the system operates near ts transmsson capacty lmt wth a reduced securty margn [1]. It may not be possble to meet the demand always and to delver all blateral and multlateral contracts due to volaton of operatng constrants such as voltage and lne power flow. In Mantanng the Integrty of the Specfcatons such cases there may be a chance of occurrence of congeston. In order to releve that congeston, n many cases cost-free means such as network reconfguraton, operaton of transformer taps and operaton of flexble alternatng current transmsson system (FACTS) devces are used [2]. In some cases t may be advantageous to releve congeston by some non-cost-free control methods, such as re-dspatch of generaton and curtalment of loads [3-5]. Snce there s a large scope of events whch can lead to transmsson system congeston, t s very mportant to manage and respond to operatng condtons n whch system voltages and/or power flow lmts are volated[6].a congeston management method proposed n ths project s based on the applcaton of Demand Response programs [7]. In ths paper programs used are TOU, CPP on IEEE-14 bus system. K. Narasmharao Dept. of Eee Gvpcoe (A), sakhapatnam. II. DEMAND RESPONSE DR s defned by Department of Energy (DOE) as:"changes n electrc usage by end-use customers from ther normal consumpton patterns n response to changes n the prce of electrcty over tme, or to ncentve payments desgned to nduce lower electrcty use at tmes of hgh wholesale market prces or when system relablty s jeopardzed". DR s classfed nto two basc categores and several subgroups [8]: 1. Incentve-based programs: - Drect Load Control (DLC) - Interruptble/curtalable servce (I/C) - Demand Bddng/Buy Back - Emergency Demand Response Program - Capacty Market Program (CAP) - Ancllary Servce Markets (A/S) 2. Tme-based programs: - Tme-of-Use (TOU) program - Real Tme Prcng (RTP) program - Crtcal Peak Prcng (CPP) Program As mentoned above, n tme based programs.e. Tme of Use (TOU), Real Tme Prcng (RTP), and Crtcal Peak Prcng programs, the electrcty prce changes wth respect to the electrcty supply cost. TOU rates establsh three perods that reflect hours when the system load s hgher (peak), moderate (off-peak), lower (valley), and charge a hgher rate durng peak hours. RTP rates vary contnuously durng the day reflectng the wholesale prce of electrcty.cpp uses real-tme prces at tmes of extreme system peak. The ncentve based programs can be classfed nto three man subgroups namely; voluntary, mandatory and market clearng programs.dlc and EDRP are voluntary programs n whch there are no penaltes for not curtalng ther consumpton. DLC refers to a program n whch system operator shuts down the customer s electrcal equpment on short notce by provdng ncentve payment or bll credt. EDRP facltate wth more ncentve payments to customers for reducng ther loads durng relablty trggered events. I/C and CAP are mandatory programs and customers who partcpated n that programs are subjected to penaltes f they do not curtal consumpton when they are called upon to do so. Customers on I/C servce rates receve a rate dscount or bll credt n exchange for agreeng to reduce load durng IJERT5IS

2 ol. 5 Issue 08, August-2016 system contngences. In CAP, customers agreed to provde prespecfed load reductons durng system contngences, and are penalzed f they do not reduce the load demand. DB and A/S are market clearng programs, where large customers are encouraged to offer load reductons at a prce at whch they are wllng to be curtaled. A/S program allows customer to bd load curtalment n electrcty market as operatng reserve. III. PROBLEM FORMULATION The costs may be defned as polynomals or as pecewse lnear functons of generator output. Generator cost functons are represented as quadratc functons MS λ P Where PG s the produced power a, b and c are cost coeffcents. OPF can be formulated n the followng form [9]: The generaton dspatch s n such a way that the above functon should be mnmsed.in Deregulated Power System, because of the congeston, the Locatonal Margnal Prces (LMP) (λ) are dfferent at varous buses.hence, the money pad by the loads (E ) s greater than the money pad to the generators (R ).e. there wll always be a Merchandsng Surplus (MS) that the ISO collects. MS Run the OPF after mplementng the DR programs usng the followng methodology, then fnd the revenue of generators, loads payment and observe the LMPs at dfferent buses. Thereafter fnd the MS. In order to manage the congeston MS should be low. TOU: In TOU, the prce s lower n valley perod, medum n off peak perod and hgh n peak perod, so that the consumers reduce ther power consumpton n peak perods and shft to other perods. In order to meet the demand n peak perods, the expensve generators are used [10]. In low and off peak perods cheaper generators are suffcent to supply the load. Based on the MS value, the congeston s compared under dfferent condtons. In order to mplement ths TOU n MATPOWER, need to follow some steps whch are mentoned below. Steps for Procedure: λ P.. L mn λ. P C P G λ P. L G G 1. Accordng to the load curve, dvde the load nto valley, off peak and peak perods. 2. Observe the varaton of prces n dfferent perods whch s low n valley, moderate n off peak and hgh n peak perods. 3. Now create congeston for that system. The prces wll be dfferent at varous buses and also hgh. 4. In order to mplement TOU n MATPOWER assume some dspatchable loads and set the margnal beneft for that loads. 5. If the prce s above that margnal beneft, the load wll be curtaled usng prce senstve load concept. 6. Then there wll be a congeston relef, whch s represented by approxmately equal and decrement of LMPs at all the buses. E E R R 7. Thereafter calculate the revenue of generators, loads payment, SR and MS. 8. Compare the above terms under normal, congested and after curtalment of load condtons. CPP: In CPP, the prce s very hgh n crtcal peak perod, n whch the duraton of the perod s less compared to TOU. Same as TOU, dvde the load nto two perods.e. n crtcal peak perod and normal perod.cpp s a dynamc prcng where the prces are set before a day under crtcal contngences. In MATPOWER due to operaton of very expensve generators the prce s very hgh n crtcal peak perods than the prce n peak perods n TOU. Steps for Procedure: 1. Accordng to the load curve, the load s dvded nto normal and crtcal peak perods. 2. Observe the varaton of prces n both perods whch s low and very hgh respectvely. 3. Now create congeston for that system. The prces wll be dfferent at varous buses and also hgh. 4. In order to mplement CPP assume some dspatchable loads and set the margnal beneft for that loads. 5. Then by usng prce senstve load concept n MATPOWER, the load wll be curtaled f the prce s above the margnal beneft. 6. Now observe the congeston relef by equal LMPs and also decrement of LMPs at all the buses. 7. Thereafter calculate the revenue of generators, loads payment, SR and MS for all the condtons. 8. Compare the above terms under normal, congested and after curtalment of load condtons. In the below pages results are projected whch are mplemented on an IEEE-14 bus systems. The tables are shown whch are the OPF results where the LMPs can be observed and the bar graphs are represented the comparson between the dfferent condtons for the two program. Numercalstudes: Fg.1. Daly load curve IJERT5IS

3 TOU: In an IEEE-14 bus system case, the network embraces two generator, three synchronous condenser and eleven load buses. There are fve sources whch can meet the load where the 1,2,3 are least expensve and 6,8 generators are more expensve. The generator cost data can be found n appendx. Accordng to TOU the prce s low for off peak perods, very low for valley perods and hgh for peak perods. Assumed 1 to 8 hrs. as valley, 9 to 18hrs as off peak and 19 to 24hrs as peak perod shown n fg1. Table 1, 2, 3 represents OPF results for valley, off peak and peak perods where we can observe that the prces are low under valley Perod, moderate n off peak perod and hgh n peak perod. The cheaper generators are unable to meet the ncreased load demand n peak perods. So more expensve generators are used n the peak perods whch are quck start to meet the load. Table-1: OPF for IEEE-14 bus system durng valley perod Table-2: OPF for IEEE-14 bus system durng off peak perod ol. 5 Issue 08, August-2016 Table-3: OPF for IEEE-14 bus system durng peak perod The results mentoned above are wthout any lne lmts. From the above tables we can observe that prce s hgh n peak perods and very low n valley perods. For that system the lmts are set n such a way of congeston s created n the lne 6-13 by 1MW. Table4 represents the OPF results for congested system where the prces are dfferent at all the buses and are hgh. In order to remove ths congeston, t s assumed that consumers at 9,10and 13 buses are partcpatng n DR program whom margnal beneft s 40$.MWh. To tackle ths hgh prces, the load at these buses are reduced to zero usng matpower dspatchable load concept and the prces at all buses come to orgnal and congeston s It removed whch s shown n table-5.it s observed that the load curtalment of the elastc loads at the partcular buses are wth respect to the margnal beneft. Table-4: OPF for IEEE-14 bus system durng congeston IJERT5IS

4 ol. 5 Issue 08, August-2016 Table-5: OPF for IEEE-14 bus system after load curtalment System Table-6: Comparson of results n dfferent condtons Cost Revenue of Generators Loads Payment SR * 0.09* * 0.00* * 0.00* MS A B C The table-6 represents the comparson made between the normal, congested and curtalment of load condtons.e. A,B,C respectvely n whch the MS value s very much hgh than the two condtons. In economcal pont of vew also, the revenue of generators, loads payment are hgh compared to others.sr s represented as the securty pont of vew whch s hgh n the case of curtalment of load. From the above comparson t s observed that due to curtalment of load t s benefcal to the customers n terms of cost and also for the utlty provders n vew of congeston. CPP: In CPP, at crtcal peak tmes the prce s very hgh.the margnal beneft s assumed as 40$/MWh. The duraton of the perod of the crtcal peak perod s very less.e.20 to 23hrs. Ths program s used where the system s under crtcal condtons. In ths case the load s ncreased, so n order to meet the load n that perod the most expensve generators comes nto the operaton whch leads to hgh prces. It dynamcally vares accordng to the load.table-7 represents the system under normal condton where the prces are normal. Table-7: OPF for IEEE-14 bus system under normal condton P G Table-8: OPF for IEEE-14 bus system durng crtcal peak Table-8 represents the system under crtcal peak condton where the load s ncreased at 4, 11, 12 buses. In order to meet that load there s more generaton dspatch from the expensve source the prce s hgh.table-9 represents the system under congested condton and the congeston s created same as TOU where the prces are very hgh especally from the buses 7 to 14. Now to mtgate ths congeston the load curtalment s used whch s shown n table-10.at 9,10 and 13 buses the load s curtaled and the congeston s removed represented by equal and reduced prces at all buses. The curtalment of the load s based on the assumed margnal beneft and the prce at that partcular bus. IJERT5IS

5 Table-9: OPF for IEEE-14 bus system wth congeston Table-10: OPF for IEEE-14 bus system after curtalment of load ($/MW h) ol. 5 Issue 08, August-2016 Table-11: Comparson of results n dfferent condtons System Cost Revenue of Generators Loads Payment SR MS A B C Graphcal Representaton: The above shown results are all the optmal power flow results where the generaton dspatch and LMPs are gven at all the buses. Those LMPs are represented n graphs. Fg2 and fg3 represents the comparson of LMP under normal, congested and load curtalment condtons when mplemented TOU and CPP. The LMPs are hgh n congested case and they decreased after curtalment of load. Partcularly, the LMPs are hgh at buses nearer to the congested area. Now by curtalment of load at the dspatchable load buses the LMP at that buses are decreased * 0.00* * 0.00* Fg.2. Comparson of LMP n TOU * 0.00* Table-11 s a comparson of system under dfferent condtons. The results clearly shows that, when the customers reduce ther load under crtcal condtons then ther loads payment s less and also they have a chance to shft ther loads from peak to off peak perods. The above table also resembles the congeston relef n terms of MS. In condton C the MS s low compared to the congested case whch s our requrement. Fg.3. Comparson of LMP n CPP IJERT5IS

6 Fg.4 and fg.5 represents the comparson of MS usng TOU and CPP programs for an IEEE-14 bus system. From the bar graphs shown below we can observe that the MS value under congested s hgh and s less when there s curtalment of load. By ths comparson we can clearly state that the congeston s removed when the consumers reduced ther load. ol. 5 Issue 08, August-2016 Appendx: Table-12: Generator Cost Data Generator a[$/hr] b[$/mwhr] c[$/mw 2 hr] REFERENCES: Fg. 4. Comparson of MS n TOU Fg.5. Comparson of MS n CPP CONCLUSIONS: In ths paper, a new model s ntroduced for congeston management usng demand response programs. It s observed that by applyng DR programs MS s reduced whch s a measure of the congeston. Wthout any nstallaton of new transmsson corrdors and usng the concept of FACTS the load s reduced usng DR programs. By usng the proposed model consumers havng the property of prce elastcty get benefted n terms of reduced loads payment by curtalment of load. [1] Chrste RD, Wollenberg BF, Wangen steen I, Transmsson n the deregulated envronment, Proc IEEE 2000;88: [2] Acharya N, Mthulanathan. N. Locatng seres FACTS devces for congeston management n deregulated electrcty markets Electr. Power Syst. Res 2007;77; [3] Tuan LA, Bhattacharya K, Daalder J. Transmsson congeston management n blateral markets: an nterruptble load aucton soluton. Electr power Syst Res 2005;74: [4] Bombard E, CarpentonE, Chcco G, Gross G. The role of load demand elastcty n congeston management and prcng. In: Power Engneerng Socety Summer Meetng, IEEE2000, vol, 4;2000, p [5] Sngh K, Padhy NP, Sharma J, Congeston management consderng hydro thermal combned operaton n a pool based electrcty market, Int J Electr Power Energy Syst 2011; 33; [6] Kumar A, Srvastava SC, Sngh SN, Congeston management n compettve power market; a bblographcal survey. Electr Power Syst Res 2005; 76; [7] Yousef, T.T. Nguye, H. Zarepour, O.P. Malk, Congeston management usng demand response and FACTS devces Accepted 5 December [8] M. Parsa Moghaddam, A. Abdollah, M. Rashdnejad,. Flexble demand response programs modelng n compettve electrcty markets, Accepted 23 February 2011, ELSEIER, Appled Energy 88 (2011) [9] Shahram Jadd, and Nahd Aslan Amol, Factorng the Prce Elastcty of Demand n the Optmal Power Flow, The 8th Internatonal Power Engneerng Conference (IPEC 2007). [10] Yudong Tang, Hongkun Song, Funan Hu, Yun Zou, Investgaton on TOU prcng prncples, 2005 IEEE/PES Transmsson and Dstrbuton Conference & Exhbton: Asa and Pacfc Dalan, Chna. IJERT5IS