A Two-Stage Multi-Criteria Decision for Distributed Wind and Solar Integration Jin, Tongdan ; Chen, Yi
Total Page:16
File Type:pdf, Size:1020Kb
A two-stage multi-criteria decision for distributed wind and solar integration Jin, Tongdan ; Chen, Yi Publication date: 2015 Document Version Peer reviewed version Link to publication in ResearchOnline Citation for published version (Harvard): Jin, T & Chen, Y 2015, 'A two-stage multi-criteria decision for distributed wind and solar integration'. General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. Take down policy If you believe that this document breaches copyright please view our takedown policy at https://edshare.gcu.ac.uk/id/eprint/5179 for details of how to contact us. Download date: 29. Apr. 2020 See discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/281716832 A Two-Stage Multi-Criteria Decision for Distributed Wind and Solar Integration CONFERENCE PAPER · JULY 2015 DOWNLOAD 1 1 AUTHOR: Yi Chen Glasgow Caledonian University 39 PUBLICATIONS 107 CITATIONS SEE PROFILE Available from: Yi Chen Retrieved on: 14 September 2015 A Two-Stage Multi-Criteria Decision for Distributed Wind and Solar Integration TongdanJin, Ph.D. Ingram School of Engineering, Texas State University Yi Chen, Ph.D. Glasgow Caledonian University, UK 1 Outline l Toward Distributed Generation l Characterizing Wind and Solar Generation l Multi-Criteria Planning Model l Numerical Experiment l Conclusion Renewable Portfolio Standards in 2040 3 US=25% EU=40% China=20-25% wind wind wind Nuclear coal Gas 3 The Rise of Distributed Power Service Distributed Power Rise of Distributed Central Power Period Age Power 1880 1910 2000 Wind and Solar Farms Onsite Generation Wind power in Scotland Wind power is Scotland's fastest growing renewable energy technology, with 2574 MW of installed capacity as of April 2011. For example: WhiteleeWind Farm is the largest on-shore wind farm in the United Kingdom with 215 Siemens and Alstomwind turbines and a total capacity of 539 MW. The Clyde Wind Farm is a 350 MW on-shore wind farm near Abington in South Lanarkshire, Scotland. The Robin RiggWind Farm is a 180 MW development completed in April 2010, is an off-shore wind farm sited on a sandbank in the SolwayFirth. https://en.wikipedia.org/wiki/Wind_power_in_Scotland#Large_wind_farms Wind power in Scotland Large wind farms in Scotland: l Black Law Wind Farm l Braes of DouneWind Farm l Clyde Wind Farm l Crystal Rig Wind Farm l Farr Wind Farm l HadyardHill Wind Farm l Robin RiggWind Farm l WhiteleeWind Farm There is further potential for expansion, especially offshore given the high average wind speeds, and a number of large offshore wind farms are planned. https://en.wikipedia.org/wiki/Wind_power_in_Scotland#Large_wind_farms Type of Distributed Generation (DG) Backup DG Onsite DG 1)Generation backup 2)Onsite generation 3)Multi-node Generation 4)MicrogridSystem Multi-Node DG Up to 299 days w/o raining 2,550-3,300 hours of sunlight/year A commercial micro-grid system in China Location: Turpan, Xinjiang Year: 2009 Capacity: 13.4 MW PV Service: 7,000 homes Micro-Grid Funded: Energy Foundation China’s Renewable http://www.efchina.org/Case-Study-en/case-201S4122ys3tem05-en Energy Program Key Notations x=decision variables representing the sizing and siting t=the maintenancetime of DG system n=number of nodes in the power system m=number of available DG equipment type l=number of link of the distribution network Dj=power demand of node j, for j=1, 2, …, n Pj(x)=power generation at node j, for j=1, 2, …, n Vj(x)=voltage at node j, for j=1, 2, …, n Ik(x)=current in link k, for k=1, 2, …, l Pr=rated wind turbine power output fw(y)=wind speed distribution S=hourly irradiance on PV, random variable Ps(S)=power output of PV h=PV efficiency A=PV area To=PV skin temperature mixed integration programming problem Modeling Variable Power Output Ø Time Series Model (.e.g. ARIMA Model) Ø DG Simulations Ø Astronomy/Physics Models(Observation) Ø Moment Methods(Mean and Variance) The models of ‘Wind’ and ‘Solar PV’ Power generation are given: Wind Power Generation 0 0 £ y < v , y > v PPr Piecewise Functions ì c s m ï 3 ï æ y ö cubic power Pw (y) = íPr ç ÷ vc £ y £ vr W) ç ÷ curve vr M ï è ø ( r ï Pr vr £ y £ vs e î w o P y density function k-1 0 vc vr vs wind speed æ k öæ y ö -( y /c)k fw (y) = ç ÷ç ÷ e è c øè c ø vr Weibull Wind Speed Distribution 3 E[P(y)] = g ò x f (y)dy + Pm (F(vs ) - F(vr )) 0.16 vc c=7, k=2 c=10, k=4 0.12 vr y f 2 2 6 2 o f 0.08 E[P (Y)] = g x f (x)dx+ Pm (F(vs ) - F(vr )) pd ò vc 0.04 Var(P(Y)) =-E[P2 (Y)](E[PY())2 0 0 4 8 12 16 Wind speed y 3 Where g = Pr / vr Solar Photovoltaics(PV) Generation Beta Distribution for Solar Irradiance 0.003 a=2, b=2, (I) 0.0025 a=4, b=4, (II) ity a=1.8, b=3, (III) s n 0.002 e a=3, b=1.8, (IV) D y 0.0015 t ili b a 0.001 b o r P 0.0005 sm 0 0 200 400 600 800 1000 Solar Irradiance (W/m2) s density function a-1 b-1 G(a + b) æ s ö æ s ö P (S) =hAS(1- 0.005(T - 25)) ç ÷ ç ÷ s o fs (s) = ç ÷ ç1- ÷ smG(a)G(b) è sm ø è sm ø Where 2 as smhA(1- 0.005(To - 25)) sm=maximum solar irradiance (W/m ) E[P (S)] = s a + b Ps(S)=power output of PV s s h=PV efficiency a b s 2h 2 A2 (1- 0.005(T - 25))2 A=PV area Var(P (S)) = s s m o s (a + b )2 (a + b +1) To=PV panel temperature s s s s Key Performance Measures of DG System q Technical Ø Energy Supply Reliability Ø Power Quality (i.e. voltage stability) Ø Line Thermal Stress q Economical Ø Return On Investment q Environmental-Social Ø Carbon Savings/Climate Change q Political Ø Renewable Portfolio Standards 12 Technical Constraints Ø Reliability (loss-of-load probability): Pr{P(x) ³ D}³1-a1 Ø Power Quality: Pr{Vmin £V(x) £Vmax}³a2 Ø Thermal Limits max Pr{I(x) £ I }³a3 13 Economic Goal Cost=Install+Operation+Carbon Credits+Maintenance fC1(x,τ)= DG (x,τ) CDG (x, τ) = Ce (x) + Co (x) + Cc (x) + Cm (x, τ) æ r(1+r)h ö m n C (x) =ç ÷ x a Pc e ç h ÷åå ij ij ij m n è(1+r) -1øi=1 j=1 a Cc (x) =ååti xijcij Pij i=1 j=1 m n m æ a n ö c ç ti (c fi Fi (t i ) + cpi Ri (t i )) ÷ Co (x) =ååxijbij Pij C (x,τ) = x m åç ti å ij ÷ i=1 j=1 i=1 j=1 ç Ri (t)dt + t fi Fi (t i ) +t pi Ri (t i ) ÷ è ò0 ø 14 Environmental and Political Goals Ø Environmental Goal: Maximize Carbon Savings m n a f2 (x,τ) =qååti xij Ai (t i )Pij i=1 j=1 Ø Political Requirements m c for j=1, 2, …, n å xij Pij £ lE[D j ] i=1 Renewable Portfolio Standards 15 A Multi-Criteria Approach to DG Planning Min: fC(x,τ)= (x,τ) Decision variables: 1 DG xij (binary) ti (positive) m n Max: a f2 (x,τ) = qååti xij Ai (t i )Pij i=1 j=1 Subject to: Pr{P(x) < D} £ a1 Pr{Vmin £ V j (x) £ Vmax } ³ a 2 for j=1, 2, …, n max Pr{I k < I k } ³ a3 for k=1, 2, …, l m c for j=1, 2, …, n å xij Pij £ lE[D j ] i=1 A Two-Stage Decision Making Stage 1: Determining x (1) Min: f1 (x) = E[Ce (x)] + E[Co (x)] + E[Cc (x)] + Z1-q Var(CDG (x)) m n (1) a Max: f2 (x) =qååti xij E[Pij ] i=1 j=1 Subject To: m ³ m + Z (s 2 +s 2 )1/ 2 P(x) D 1-a1 P(x) D V - Z s £ m £V + Z s min (1-a2 )/ 2 V j (x) V j (x) max (1-a2 )/ 2 V j (x) m £ I max - Z s Ik (x) k a3 Ik (x) m c å xij Pij £ lE[D j ] i=1 Stage 2: Determining t f (2) (τ;x) = f (1) (x) + E[C (τ;x)] Min: 1 1 m m æ n ö f (2) (τ;x) = q çt a A (t ) x E[P ]÷ Max: 2 åç i i i å ij ij ÷ i=1 è j=1 ø Subject To: t i > 0 Numerical Experiment: A13-Node Network Input Data WT and PV Options (m=5) 1)The mean and the standard deviation of Dj i DER Capacity (MW) 2)Wind speed and solar irradiance distributions 1 WT1 1.0 3)Wind turbine (WT) power curve 2 WT2 1.5 3 WT3 2.0 4)Costs associated with WT and PV maintenance 4 PV1 0.5 5)Lifetime distribution of WT and PV units 5 PV2 1.0 6) a1, a2, a3, and l Testing on 13-Node Network (n=13) Output Data link 12 13 1)Where to place the WT/PV? 2 11 node 2 1 10 D x 13 D2 D 11 2)The size of WT/PV units 3 1 11 3 4 8 D D 12 t 3)The maintenance time of WT/PV 3 1 5 5 9 4 9 D12 f1 4)Annualized system cost 6 7 6 D 10 5 D9 D4 D f 5)Carbonsavings estimation 6 12 7 8 D10 D7 D8 Pareto Solution in Stage 1 Pareto Solution in Stage 1 2.8 2.6 ) 2.4 $ 8 n i 2.2 n o i 7 2.0 6 5 Mill B ( 1.8 4 t s 3 1.6 2 Co 1 Pareto Frontier 1.4 A 1.2 2.5 3 3.5 4 4.5 5 5.5 Carbon Savings (10,000 tons) Selected Pareto Solutions in Stage 2 Solution No.