Energy Management, Voltage and Frequency Control for Smart Grids - A Technology Providers View

Transcription

1 Michael Mezger Siemens Corporae Technology Munich Energy Managemen, Volage and Frequency Conrol for Smar Grids - A Technology Providers View 2014 IEEE Muli-Conference on Sysems and Conrol 8-10 Ocober, Anibes/Nice, France Resriced Siemens AG All righs reserved

2 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion Page 2 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

3 Siemens has a long radiion in grid auomaion, bu new hings are happening Siemens has a long radiion in grid auomaion, bu new hings are happening Folie 3 Page 3 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

4 Significan changes in energy sysems From cenralized, unidirecional grid Page 4 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

5 Significan changes in energy sysems o disribued energy and bidirecional energy balancing Hydrogen Sorage Diesel Generaor Biogas CHP Offshore Wind Parks Small Indusrial Gas Turbine Privae Solar Pumped Sorage Power Plan Sorage Soluions Sorage Soluions Elecrical Vehicles Smar Sree Lighing Large Scale PV Plan Unidirecional Power Flow Bidirecional Power Flow Page 5 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

6 Smar Grid / Generaion Definiions Microgrid Microgrids are geographically resriced areas conaining and conrolling generaion asses, sorage and consumpion. A Microgrid does have defined poin(s) of common coupling Microgrids manage he asses in a decenralized and smar way in order o generae he maximum cusomer benefis of he asses. Microgrids oday show an increasing share of flucuaing renewable generaion May be pure elecrical or opionally inegrae heaing and cooling equipmen Island Grid 1) (sub-se of Microgrid) Island grids are Microgrids, which can run independenly and which are able o mainain volage and frequency (for AC) Island grid migh be able o swich beween off-grid and on-grid mode Island grid migh be are able o black-sar Nano Grid (sub-se of Microgrid) Nanogrids are small Microgrids much smaller han 1 MW insalled power Power neworks in a building or facory, ypically low volage Virual Power Plan (Demand Response) VPP and demand response describes an economic aggregaion of generaion and consumpion asses in order o rade he produced energy (or shedded load). VPP are no able o run grids VPPs generaion asses can be in broader geographical region (e.g. naionwide) VPP asses are always on-grid 1) Island grid also referred o as island mode Microgrids Page 6 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

7 There is no one global Smar Grid Regional drivers are differen: some examples Aging infrasrucure Non-echnical losses Renewables Inegraion Elecrificaion of Megaciies Disribued Generaion Page 7 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

8 Key challenges drive renewable inegraion Trends Cusomer challenges Renewable generaion in disribuion grids Increasing elecrical loads in LV disribuion grids Aging and/or weak infrasrucure Overload of disribuion grids due o flucuaing renewable in-feed E.g. share of world renewable generaion o riple from 4% o 13% in ) High cos for inegraion of renewable generaion hrough grid exension Limied ransparency on disribuion grid Disribuion grids are no designed for bidirecional energy Source: 1) Energy Trends Sudy Page 9 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

9 Changing in-feed paerns challenge exising grid infrasrucures Week burden of a ransformer saion in he rural area he LEW-Vereilnez GmbH 2003 and oday Load in kw Load profile 2003 Load profile oday Disribuion grid changes from load o source 12:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 12:00 0:00 Source: LEW Page 10 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

10 Problems in LV Grids due o increasing PV in-feed Page 11 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

11 Problems in LV Grids due o increasing PV in-feed Page 12 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

12 Problems in LV Grids due o increasing PV in-feed Page 13 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

13 Fuure Siuaion for Germany will be challenging: Load and Generaion will differ in ime and locaion Day 124, Year 2050: 73 GW PV according Leisudie 2011 A (BMU), Spaial Disribuion like 2012 Page 15 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

14 Key challenges drive implemenaion of he Microgrid Trends Cos opimized operaion of remoe grids Power reliabiliy and independence Load balancing Cusomer challenges High cos of operaions for remoe grids based on diesel generaion. Inegraion & balance of he high proporion of cos-efficien renewable generaion in remoe grids. Power availabiliy and reliabiliy. Opimized operaion hrough balancing generaion, consumpion and sorage. Independence. Inegraion of renewables Page 16 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

15 Smar Inverer are becoming a key componen for our Energy Sysem Micro-Grid- Conroller Bi-direcional elecric vehicle charging saion power informaion Inverer Inverer Inverer Transformer Swiching -Saion Smar Meer Inverer Inverer Inverer IT inegraion Page 17 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

16 MV LOADs LV LOADs Sysem design for Microgrids An example seup Demand: ~4 MWp PV Generaor: ~5 MWp Baery: ~3 MW / 2 MWh Diesel Generaor(s): 2 x 2 MWp Microgrid Managemen (SICAM (small) & Specrum Power (large) µg Manager) incl. Weaher Forecas-Based Opimizaion Auomaed Generaion Conrol (frequency, volage and cos phi conrol) Remoe Monioring Inerface G Diesel Gen Se(s) ~ PV Plan = PV Inverer(s) Baery sorage = ~ Baery Inverer(s) + Baery Managemen local (islanded) MV disribuion grid Page 18 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

17 Key challenges drive implemenaion of Demand Response & Virual Power Plans Trends Cusomer challenges Generaion & nework bolenecks Increasing peak load prices Increasing disribued & renewable generaion Rising consumpion Generaion & nework capaciy bolenecks: E.g. California, US. Dispach load as mos economic power supply: Avoidance of generaion and nework bolenecks and high peak load prices. Increased grid sabiliy hrough emergency load shed & selecive load dispach. New marke opporuniies for disribued energy resources. Page 20 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

18 The challenge decenralized generaion: High complexiy of engineering, algorihms and daa models DENMARK 1980 AND 2005 Source: A*STAR Energy Technology R&D Program, Prof Ho Hiang Kwee Program Direcor, Page 22 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

19 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen Opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion Page 24 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

20 Self-Organizing Energy Auomaion Sysems coordinaing smar componens wihin he grid MULTI-AGENT ARCHITECTURE FOR POWER GRIDS (SCHEMATIC) *) Design Principles: *) *) Smar componens. Use plug and play for engineering. *) Nework Transpor Agen Coordinaion of local algorihms whenever necessary *) Personal Energy Agen Balance Maser *) Böse, C.; Hoffmann, C; Kern, C.; Mezger M.: New Principles of Operaing Elecrical Disribuion Neworks wih a high Degree of Decenralized Generaion, 20h Inernaional Conference on Elecriciy Disribuion, Prague (2009). Area Adminisraor Elecriciy Police Page 25 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved *) *) *) Personal Energy Agen

21 local conrol coordinaion opimizaion The modular srucure enables daa privacy and spli of ownership Three Layer Conrol Archiecure price negoiaion or signal or direc conrol Balance Maser long erm energy balancer grid load indicaor P, Q se poins droops Area Admin shor erm balancer volage conrol NTA* load flow esimaion, diagnosis Measuremens U,I,P,Q PEA** local fas fas fas conrol conrol power frequency, power qualiy qualiy Generaion, Power Generaion // Consumpion, Consumpion Sorage Grid * Nework Transpor Agen ** Personal Energy Agen power flow Informaion flow Page 26 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

22 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen Opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion Page 27 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

23 Several Siemens divisions provide essenial componens and applicaions Divisions and Business Uni wih relevan porfolio smar energy / grid Digial Facory & Process Indusries Building Technology Indusrial Load and Generaion Managemen Building Energy Managemen Energy Performance Conracing Demand response Virual Power Plan Energy Managemen - Energy Auomaion - Smar Grid Soluions & Services Disribuion Auomaion Microgrid Disribued Generaion Elecric Vehicle charging Power and Gas, Wind and Renewables Smar Generaion Soluions and Hybrid Power Plans Energy Managemen - Transformers Conrollable Transformers and HVDC sysems Energy Managemen MV and LV Primary Equipmen: -Sorage sysems -Swiching - LV equipmen Drives Smar Converer Corporae Technology drives he overall energy sysem design Page 28 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

24 In a comprehensive approach, we consider he whole energy supply chain for elecric power, hea/cold and gas Waer Producion Gas Supplier Offshore Wind Park Onshore Wind Park Gas Transpor Waer Transpor Power o gas (liquid, chemical) Pumped Hydro Gas Saion Sorage Onshore Wind Biomass Indusry Pumping Saion C(C)HP** Disric heaing cooling Thermal Sorage Gas Disribuion Waer Tanks Waer Disribuion Heaer, Chiller Hea Saion Sorage Soluions Smar Building w/ hea pump & hermal sorage Disric heaing cooling Microgrid Fuel cell Privae Wind or Solar Microgrid Conroller Engine Generaor wih C(C)HP** Elecrical Vehicle Infrasrucure Nework Conrol Cener Large Scale PV Plan Conrollable LV ransformer Fossil Power Plan Power Saion Power Saion Power Saion Bulk Generaion Transmission Disribuion & DER* Cusomer and Indusry Nanogrid wihin Building (e.g. DC grid in Daa Cener) Energy Managemen Sysem Smar Meer Power AC Hea / Cold (Producion Plan) Smar Energy Agen Power DC Gas * Disribued Energy Resources Mulimodal Access Poin Waer supply Building Energy Managemen Sysem ** Combined (Cooling) Hea & Power Page 29 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

25 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen Opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion P. Wolfrum, M. Kauz, J. Schäfer, Smar operaion of CHP unis Power Plans and Power Sysems Conrol 8 (1), 61-66, 2012 Page 32 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

26 Energy Agens managing Componens wih Hea and Elecriciy Sorage Concep of operaion manager for primary equipmen Complee Sysem: Subsysem under examinaion: Comp. Operaion Manager EP PEA PEA Muli-agen approach: PEA LEX Reducion of effor & complexiy a cenralized conrol insances hrough disribued inelligence Responsibiliy of operaion sraegy & conroller parameerizaion says wih componen owner. AA PEA Page 33 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

27 e Pyranomeer Evacuaed Tube Collecor Model based conrol requires reduced model TDWT complexiy 6 bar SC2 FIR SC1 Hea ransfer fluid: TYFOCOR-LS SC3 SC4 STC2 STC1 Hea ransfer fluid: waer FIR STC1 3 bar STC3 STC4 ST HC4 FIR HC2b HC3 PIR Hea ransfer fluid: waer 04 HC7b HC8b RC1 RC2 3 bar SC1 SC6 SC5 HTC4 Saring from hermodynamic layou Emergency Hea Dump HTC5 FIR HC2 FIR FIR CC8 CC7 CC2 CC4 CTC4 CTC1 Cold ransfer fluid: waer and sysem heoreic relaions of he CHP uni Hea ransfer fluid: waer FIR CC1 CC3 Hea Loss FIR HTC1 FIR HC1 CC1 CC2 CTC1 CTC2 CTC3 HC8 Cooling Tes Cener Air (aspiraed by inernal fan) PIR X Propane E3 E1 FIR E1 E2 Exhaus Ne ETC2 CHP-Uni ETC1 HTC1 HTC6 Heaing Tes Cener FIR ETC1 3 bar HTC2 ETC3 FIR ETC2 ETC4 ET HTC3 HC2 HC1 HC5 HC6 HC7 Energy Inpu CWC1 Energy Inpu TDC CWC2 Hea ransfer fluid: waer FIR FIR CWC1 CWC2 Elecriciy Backup CC5 Heaer CWC3 (opional) CHP - Uni CWC6 CWC5 CC6 CWC4 Hea Loss Hea Load SIEMENS AG Corporae Technology Developmen Cener Decenralized Poly-Generaion Größ Zeichnungsnummer Rev. A Hea Sorage (TES) Bla 1 von 1 CT T EEE Schäfer PTP Hea Loss reduce model complexiy whenever possible Thermodynamic modeling of cooling sysem is very complex Experimenally deermine hermal mali/boni incurred a swiching Page 34 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

28 SOC [-] Behavior of he CHP Model wih Thermal Correcion 1,2 Sae of Charge of TES* 1,0 0,8 0,6 0,4 0,2 0,0 Measured Prediced w/o bonus/malus ime [h] Accurae predicion of hermal dynamics! * Thermal Energy Sorage Page 38 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

29 Use model based CHP Operaion Manager for neiher hea nor power led bu cos opimal operaion Funcionaliy Inpus, oupus, objecive funcion Elecriciy prices Opimized schedule Thermal loads Componen Models (Gas engine + Thermal Energy Sorage) User Specs (weighing facors + consrains) Opimal Conrol (Mixed Ineger Programming) Objecive funcion: Fuel and CO2 emission allowances Revenue C OM C CHP,CAPEX C BH,CAPEX C CHP,GAS&CO2 C BH,GAS&CO2 C CHP,o&m C BH,o& m R Sar penaly Operaion and Mainenance Page 39 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

30 Sae-Transiion Model Sae s: v Є [0,1,,vmax] : idle ime of CHP. v=i means an idle ime of i d q Є [0,dQ,,1] : normalized SOC of hermal sorage, finely discreized. Acion u: uє [0, u min, u min +du,,1] : CHP aciviy (normalized o max power), sufficienly discreized. u has o be such ha SOC never falls below 0 or above maximal fill level (see dynamics below) Page 42 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

31 Page 45 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved )),, ( ( ),, ( arg min ), ( )),, ( ( ),, ( ), ( ) ( ) : ( 1 * 1 ), ( min w x U u N N N N w u s f J w u s g w s u w u s f J w u s g w s J s g s J Dynamic Programming Now use he sae-ransiion model o find an opimal CHP conrol sraegy using Bellman s principle of opimaliy Goal: Minimize oal cos Define opimal cos-o-go Then calculae opimal cos and policy backwards (using opimaliy principle) 1 0 ),, ( N N w u s g g G w is assumed deerminisic and known here (exernal energy prices). J ),, ( 1 w u s f s

32 Sae of Charge [-] Elecric Power [kw] Thermal Load [kw] Eleciciy Price [ ] Resuls Behavior for Arificial Show Case prognosis hea-led operaion power-led operaion operaion manager ime [h] Page 48 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

33 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen Opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion Szabo, A; Buhl, M.; Mezger, M.: Volage Conrol for Disribued Energy Resources - The Value of Coordinaion, IEEE Inernaional Energy Conference (ENERGYCON), 2014, Dubrovnik, Croaia Page 53 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

34 local conrol coordinaion opimizaion The modular srucure enables daa privacy and spli of ownership Three Layer Conrol Archiecure price negoiaion or signal or direc conrol Balance Maser long erm energy balancer grid load indicaor P, Q se poins droops Area Admin shor erm balancer volage conrol NTA* load flow esimaion, diagnosis Measuremens U,I,P,Q PEA** local fas fas fas conrol conrol power frequency, power qualiy qualiy Generaion, Power Generaion // Consumpion, Consumpion Sorage Grid * Nework Transpor Agen ** Personal Energy Agen power flow Informaion flow Page 54 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

35 The Pilo Projec IRENE (Disribuion Nework near Munich) hp:// Medium volage grid 20 km Tes area Measuremens a each subsaion Measuremens a seleced house connecions/ generaors / e- Cars 87 secondary subsaions 32 Srompioniere wih elecric vehicles 3MW peak and 0,6MW minimum load 6MW PV, 2MW Wind, 1MW Biogas Asymmery a he ransformer an he end of he feeder Volage conrol via PV inverer Page 56 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

36 For masering he complexiy, local inelligence needed New acuaor in he disribuion grid Conrollable subsaion Transformer (RONT) CHP Uni + Warm waer ank Charging saion for E-Cars Renewable generaion Baery sysems Page 58 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

37 Conrollable subsaion Transformer (RONT) help o inegrae PV plans wihou grid exension Acion principle RONT Page 59 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

38 Conrollable subsaion Transformer (RONT) help o inegrae PV plans wihou grid exension Acion principle RONT Page 60 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

39 Conrol Sraegy Basic principle - As much local inelligence as possible. U 1, U 2, a Q a U 1, U 2, a Q a If U >U D, ap = ap -1 If U <U U, ap = ap +1 U U 1 U 2 U U U 1 U 2 U RONT Grid U U 1, U 2, a Q a U 1, U 2, a Q a U D, U P U U 1 U 2 U U U 1 U 2 U AA Prognosis of power NTA Conrol cener Auomaed configuraion of he local conrollers Opimizaion of he local conrollers: Droop curves vs. I conrollers wih lower bound Page 61 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

40 Conrol of he inverer conneced generaors (PV) Conrol sraegies: No conrol (now) fixed Power Facor (Q = func(p)) droop conrol (Q = k * (U U 0 )) Inegral conrol: Advanages of I-conroller: - Predicable dynamic behavior - Easy and sable parameerizaion - Reducion of wase of reacive power - Simple and opimal acive power curailmen (if needed and allowed) u u se q q max p p PV Page 64 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

41 Experimenal resuls of he PV conrol Variaion of he sepoin o impose he acivaion of conrol Firs conrol wih Q, disribued inverers help each oher in he conrol When Q reaches is limi, hen also acive power is curailed Page 65 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

42 Energy Managemen, Volage and Frequency Conrol Agenda Use Cases & Challenges Self-organizing Grids Self-organizing Grids wihin Siemens Energy Managemen Opimal CHP Operaion Volage Conrol acive Disribuion Grids Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion Mezger, M.; Muenz, U.: Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion, 19h IFAC World Congress, 2014, Cape Town, Souh Africa. Muenz, U.; Romeres, D.: Region of Aracion of Power Sysems, 4h IFAC Workshop on Disribued Esimaion and Conrol in Neworked Sysems, 2013, Koblenz, Germany Page 66 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

43 local conrol coordinaion opimizaion The modular srucure enables daa privacy and spli of ownership Three Layer Conrol Archiecure price negoiaion or signal or direc conrol Balance Maser long erm energy balancer grid load indicaor P, Q se poins droops Area Admin shor erm balancer volage conrol NTA* load flow esimaion, diagnosis Measuremens U,I,P,Q PEA** local fas fas fas conrol conrol power frequency, power qualiy qualiy Generaion, Power Generaion // Consumpion, Consumpion Sorage Grid * Nework Transpor Agen ** Personal Energy Agen power flow Informaion flow Page 67 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

44 Power sysems wih high renewable peneraion are more volaile han convenional power grids. Moivaion Convenional power sysems Small number of large generaors Power conversion by synchronous generaors Large ineria damps frequency oscillaions PV passive load D DG G passive load Wind G Baery Inner conroller acs hrough roaing mass slow dynamics ΔI ΔI ΔI ΔI ΔI ΔI ΔI ΔI ΔI ΔI ΔI ΔI Power sysems wih high renew. peneraion Large number of small generaors Power conversion by power converers ΔI ΔI ΔI ΔI Less damped frequency oscillaions because of lower ineria Inner conroller acs hrough power converer fas dynamics passive load M M ASM Page 68 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

45 Power Sysem Sabiliy for Microgrids is a key elemen Power Sysem Sabiliy and Conrol in Microgrids Today Disribuion grid and building power supply permanenly coupled o public grid Disribued energy resources are reaed as negaive loads (no auxiliary services) Fuure Disribuion grids and building power wih disribued energy resources can provide UPS* in he case of blackous Disribued energy resources conribued o volage and frequency sabiliy and proecion Real island grids are dominaed by Diesel generaor ses Diesel generaor ses form he grid in oday's island grids Diesel-Gen-Ses provide necessary shor circui power for proecion The maximum share of renewable sources, hus poenial Diesel off-seing, is limied Compleely inverer driven or inverer dominaed elecrical islands are possible Grid forming can be done by solar of sorage inverer Smar inverer provide proecion funcionaliy Diesel off-seing only limied o availabiliy of renewable sources. * Unineruped Power Supply Page 69 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

46 elecric hardware conrolleable componens power managemen Primary conrol is key level for sabiliy Fuure power sysem conrol archiecure Conrol levels Cenral cos-opimal power scheduling Cenral coordinaion and compensaion of primary conrollers Disribued primary conrollers on componen hardware cos-opimal power scheduling Coordinaion and compensaion of U/f conrol eriary conrol secondary conrol Inner conrol loops Feaures of his archiecure Resembles archiecure of convenional power grids Compaible wih sae-of-he-ar componens U/f conr. inner conr. Diesel gense U/f conr. inner conr. phoovolaik U/f conr. inner conr. wind U/f conr. inner conr. sorage U/f conr. inner conr. conr. load primary conrol No communicaion required for firs reacion on load/generaion variaions, fauls, ec. power grid passive load passive load acive load passive load passive load Page 70 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

47 Primary droop conrol = decenralized P-conroller Fundamenals of primary droop conrol Primary droop conrol Decenralized proporional conrollers for volage ampliude and frequency or equivalenly acive and reacive power Developed for ransmission grids wih R/X 0 -k p1 -k p2 f i -f 0 p i -p i0 Tuning rules Small seady-sae offse requires large gains Sabiliy and robusness require small gains Primary reserve sharing requires -k q1 -k q2 u i -u i0 q i -q i0 Page 71 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

48 Join volage and phase sabiliy is coupled consensus problem. Power sysem modelling Power converer model: Droops & volage low pass filer Volage a bus i: Power injecion a bus i: Droop gains parameers: Droop sepoins: PT1 filer cu-off frequency: Power grid model: AC power flow model Volage a bus i: Power injecion a bus i: Impedance beween bus i and j: Se of buses adjacen o bus i: Power sysem model for R/X = 0, i.e. : Coupled consensus problem Page 72 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

49 A lower bound on he region of aracion is deermined by conracion argumens. Power sysem sabiliy and reserve Main seps Cener seady sae a origin: Define smalles diamond-shaped se ha conains all saes Show ha conracs o a single poin for Main resul Given an inverer based power sysem and assume ha a seady sae exiss. This seady sae is asympoically aracing for all iniial condiions in if 3 derived analyically given inequaliies hold. Main observaions sandard recangular conracion argumen does no work because of coupling The 3 inequaliies can always be saisfied for sufficienly small ses Mezger, M.; Muenz, U.: Volage and Angle Sabiliy Reserve of Power Sysems wih Renewable Generaion, 19h IFAC World Congress, 2014, Cape Town, Souh Africa. Page 74 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

50 elecric hardware conrolleable componens power managemen New conrol mehods are needed for he design of modern power sysems. Summary and oulook Summary Analyical sabiliy condiion for join volage and phase angle sabiliy cos-opimal power scheduling eriary conrol Analyical lower bound on region of aracion Similar resuls for power grid wih idenical R/X raio and droop roaion Coordinaion and compensaion of U/f conrol secondary conrol Open iems and oulook Exension for non-zero R/X raios no ye published Derive design condiions U/f conr. inner conr. Diesel gense U/f conr. inner conr. phoovolaik U/f conr. inner conr. wind U/f conr. inner conr. sorage U/f conr. inner conr. conr. load primary conrol Exension for higher-order dynamics require differen proof echniques 1) power grid passive load passive load acive load passive load passive load 1) Münz and Romeres: Region of Aracion of Power Sysems. NecSys, Page 76 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

51 Power Sysem Sabiliy for Transmission Sysems is of increasing Ineres Power Sysem Sabiliy and Conrol in Transmission Sysems Today Dominaing Synchronous Generaors Predicable and smooh varying power flows Well acceped grid codes defining desired behavior of in-feed and load Limied number of wors case scenarios Fuure High share of renewable generaion Highly flucuaing power flows Grid codes and faul ride hrough condiions have o be (re-)defined for converer dominaed grids Exponenial growh of wors case scenarios Power sysem analysis Conroller design based on experience and heurisics Non-linear ime domain simulaions of wors case scenarios Linear sabiliy analysis in frequency domain Power sysem analysis Model based conroller design and auomaic (re-)configuraion Analyical condiions for sabiliy analysis Non-linear mehods in he ime domain Page 77 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

52 Model Transmission Sysem and Generaors Coupled Swing Equaion: Page 79 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

53 Model Sable Condiion: Page 81 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

54 Lyapunov Analyzis Kineic Energy: Poenial Energy: Lyapunov Funcion Page 85 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

55 Region of Aracion Explici Condiions Region of Aracion wih Quadraic Bound and Weighed 2-norm Region of Aracion wih Quadraic Bound and Non Weighed 2-norm Region of Aracion wih Quadraic Bound and Non Weighed -norm Münz and Romeres: Region of Aracion of Power Sysems. NecSys, Page 94 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

56 Innovaive conrol sysems are crucial for power sysem sabiliy and performance in Smar Grids Key Messages The energy world is changing rapidly and is a exciing place for a conrol engineer o be. There are differen drivers for Smar Grids in he world, bu he renewables are dominaing. The change is no limied o he elecric disribuion sysems, bu affecs also power ransmission and even he hea/cold, gas and waer infrasrucure New power elecronic acuaors and renewable generaion will help o opimize efficiency, susainabiliy & securiy of supply. Scalable soluions for energy managemen, volage and frequency conrol are of growing ineres Inside real world applicaions here is he need for pleny of mahemaics and conrol heory Page 96 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved

57 Thank you for your aenion! Conac Daa Dr. Michael Mezger Principal Key Exper Power Managemen Corporae Technology / CT RTC PET Oo-Hahn-Ring 6, Munich, GERMANY Phone: +49 (89) Fax: +49 (89) Mobile: +49 (1520) michael.mezger@siemens.com hp:// Page 97 Ocober 2014 Corporae Technology unresriced Siemens AG All righs reserved