Smart Electricity Systems - Demand Response - PDF

02RUK - Smart Electricity Systems USERS’ PARTICIPATION AND DEMAND RESPONSE Prof. Gianfranco Chicco Politecnico di Torino Dipartimento Energia “Galileo Ferraris” © Copyright Gianfranco...

02RUK - Smart Electricity Systems USERS’ PARTICIPATION AND DEMAND RESPONSE Prof. Gianfranco Chicco Politecnico di Torino Dipartimento Energia “Galileo Ferraris” © Copyright Gianfranco Chicco, 2012-2022 GENERAL CONCEPTS ON LOAD MANAGEMENT © Copyright Gianfranco Chicco, 2012-2020  Demand Side Management n The techniques to manage the user’s load were already studied in the Eighties, with research programmes launched in the U.S. under the name Demand Side Management (DSM) n The DSM objectives were to obtain variable evolutions of the electrical load during time, by applying some basic principles: q peak shaving (or clipping) q valley filling q load shifting q strategic load reduction (conservation) q strategic load growth q flexible load shape © Copyright Gianfranco Chicco, 2012-2020  2. RESEARCH PROBLEM OVERVIEW DSM basic principles _______________________________________________________________________________________________________ C.W. Gellings, J.H. Chamberlin, Demand-side management: Concepts & Methods, The Fairmont Press, Inc., 1993. Figure 2.3. Load control strategies (Gellings, 1993). © Copyright Gianfranco Chicco, 2012-2020  Evolution of the tariff structures n In parallel with the DSM studies, the tariff structures evolved with the creation of time-dependent tariffs: n Time-Of-Use (TOU) rates: q Based on the partitioning of the year into periods with different tariff rates (e.g., multi-hour tariffs with peak hours, intermediate hours, and low-load hours) q The main aim of these tariffs is to convince the users to move their consumption to the low-load hours q The tariff rates are predefined and are kept constant for a relatively long period of time (e.g., one year) n Real-Time Pricing (RTP): q Evolution of the TOU concept with variation of the tariff rates during time (e.g., at each hour) in a way defined closer to the real time (e.g., one day before) n Spot Pricing (SP): q Situation in which the tariff rates are defined immediately before the instant of consumption (conceptually with no lower limit, e.g., minute-by-minute) © Copyright Gianfranco Chicco, 2012-2020  Load management today n Nowadays, the basic aspects of load management correspond to different degrees of customer participation to the evolution of the electrical load, and can be seen from different perspectives n Load Shedding (LS) q activity aimed at cutting out loads, driven by the System Operator to deal with emergency conditions of the system; priorities can be given to different customers, with no customer participation n Interruptible Load Management (ILM) q activity aimed at cutting out loads following specific programmes agreed between System Operator and customers, to be applied at specific conditions and subject to remuneration as a specific service © Copyright Gianfranco Chicco, 2005-2020  Load management today n Real-Time Pricing (RTP) q variation of the electricity rates during time (e.g., at each hour, or less), communicated in advance by the provider (e.g., the day before, or less) to the consumers n Demand Response (DR) q voluntary reduction of the electricity use in response to price incentives, aimed at improving the effectiveness of the electricity supply and reducing demand peaks n Direct Load Control (DLC) q activities aimed at changing the shape of the load pattern according to the willingness to accept modifications of the supply characteristics with respect to standard or pre-defined conditions; can be subject to automatic control © Copyright Gianfranco Chicco, 2005-2020  Controllable loads n Temperature-based loads (electrical heating and cooling devices, air conditioners, refrigerators): q concur in forming the peak load q are typically driven by thermostats q the cycling operation of the thermostats is easily controllable q no need for interrupting the supply, only adjustments q the limits of application of load control has to be specified (e.g., minimum or maximum temperature allowed in the space) q the customers’ acceptance of this type of control is generally good n Lighting: q supply of lighting devices can be arranged into groups q some groups may be seen with lower priority q a minimum lighting effect has to be satisfied n Low-priority appliances: q can be subject to intermittent interruption without significantly affecting their performance, or can be supplied from dedicated paths and subject to possible curtailment in case of need © Copyright Gianfranco Chicco, 2005-2020  Deferrable and Curtailable Loads § Deferrable load may be classified into: à Non-flexible load such as Washing Machine (WM), with a predefined pattern that cannot be altered during operation time à Flexible load such as Plug-In Hybrid Electric Vehicle (PHEV) § Curtailable load may be classified into: à Partially curtailable load such as Air Conditioning (AC): their power consumption can be controlled according to its temperature set point à Fully curtailable, according to the consumer priorities: these appliances can be switched off without the need of re-turn them on later (for example, light bulbs) S. Altaher, P. Mancarella, and J. Mutale, Automated Demand Response from Home Energy Management System under Dynamic Pricing and Power and Comfort Constraints, IEEE Transactions on Smart Grid, Vol. 6, No. 4, July 2015, 1874–1883.  Deferrable and Curtailable Loads Appliance Classifications Deferrable Appliances Curtailable Appliances Non-Flexible Flexible Partially Curtailed Fully Curtailed (e.g., WM) (e.g., PHEV) (e.g., AC) (e.g., Light bulb) Consumer Preferences Waiting time End time, & Based on Consumed Set points Periorities preference starting energy time Optimization Engine Find optimum set points Power consumption, Starting time Curtailed Power Status (on/off) starting & end time Modify consumer profile S. Altaher, P. Mancarella, and J. Mutale, Automated Demand Response from Home Energy Management System under Dynamic Pricing and Power and Comfort Constraints, IEEE Transactions on Smart Grid, Vol. 6, No. 4, July 2015, 1874–1883. DEMAND RESPONSE © Copyright Gianfranco Chicco, 2012-2020  Definitions and categorisations § The concept of demand response may be seen as the variation in the user’s consumption with respect to their consumption in normal conditions, in response to specific programmes § A distinction refers to the type of programme: Ø based on incentives Ø based on price signals § The programmes based on incentives may be: Ø classical, in which the users receive revenues referring to their participation Ø market-based, in which the users receive revenues for their performance referring to particular types of services M.H. Albadi, E.F. El-Saadany, A summary of demand response in electricity markets, Electric Power Systems Research 78 (2008) 1989–1996 © Copyright Gianfranco Chicco, 2012-2020  Incentive-based programmes § Some classical programmes with incentives: Ø Load control (e.g., by switching off for a short time devices such as air conditioners or water heaters) Ø Interruptible/Curtailable load programmes (the user reduces the load by a pre-determined amount) § Some market-based programmes: Ø Offers sent by the users (demand bidding) to reduce the load in various types of markets Ø Emergency programmes, with load reductions in case of need (e.g., during a heat wave) Ø Capacity markets, with load reduction to provide reserves Ø Other ancillary services, with load reduction (or increase) offered or delivered in response to real- time price signals M.H. Albadi, E.F. El-Saadany, A summary of demand response in electricity markets, Electric Power Systems Research 78 (2008) 1989–1996 © Copyright Gianfranco Chicco, 2012-2020  Price-based programmes § Some programmes based on prices: Ø Time of Use (TOU): prices defined on the basis of the time periods (evolution of the multi-hour tariffs) Ø Critical Peak Pricing (CPP): specific prices defined for a limited number of days in the year (e.g., when a heat wave or cool wave is expected), known in advance (more than one day), with TOU tariffs in normal days Ø Extreme Day CPP (ED): similar to the CPP, with further specific prices defined for “extreme” days Ø Extreme Day Pricing (EDP): prices for extreme days not programmed and defined in the short term (e.g., the day before) Ø Real Time Pricing (RTP): price variable during time, defined on the basis of the cost of electricity, known on a day-ahead or hour-ahead basis M.H. Albadi, E.F. El-Saadany, A summary of demand response in electricity markets, Electric Power Systems Research 78 (2008) 1989–1996 © Copyright Gianfranco Chicco, 2012-2020  Costs and benefits for Demand Response Example of partitioning of costs and benefits M.H. Albadi, E.F. El-Saadany, A summary of demand response in electricity markets, Electric Power Systems Research 78 (2008) 1989–1996 © Copyright Gianfranco Chicco, 2012-2020  Benefits for the system n Aggregation Seviced KEY for succes The users’ participation may be useful to prevent potential problems for the system, for examplecan Aggregator if anbe excessive consumption third party, utility,isor expected retail supplier Aggr Resp source J. Stromback, Status of Demand Response in Europe, Smart Energy Demand Coalition, EPFL Lausanne, Switzerland, 2015 The aggregator collects multiple © Copyright Gianfranco Chicco, 2012-2020  Benefits for the system n An aggregator could be needed to manage the situation in a coordinated way n The aggregator may be the distributor, the retailer, or another entity n The aggregator collects the availability of many users with several components which consumption may be reduced or shifted in time in a flexible way n The demand side may participate in the enhancement of the operating conditions in different time horizons, also linked to the markets n The variations that correspond to the users’ action are calculated with respect to a reference (baseline) case that has to be identified, for example in a conventional way n An accurate measurement system is needed to verify that the service requested to the user has been actually delivered by the user n The incentives have to be tangible to attract the users J. Stromback, Status of Demand Response in Europe, Smart Energy Demand Coalition, EPFL Lausanne, Switzerland, 2015 © Copyright Gianfranco Chicco, 2012-2020  Demand Response Economic Energy Dispatch Benefits for the system 8/17/2009: Via phased DR, 75MW of expensive generation avoided J. Stromback, Status of Demand Response in Europe, Smart Energy Demand Coalition, EPFL Lausanne, Switzerland, 2015  Demand Response in Europe n The European Energy Efficiency Directive (2012/27/EU) contains various references to the use of Demand Response techniques, e.g.: q The article 15.4 requires to the Member States to remove the incentives that may reduce the participation of Demand Response actions to the system services, and to enhance the users’ participation to energy efficiency and Demand Response q The article 15.8 indicates that the National Authorities (e.g., ARERA in Italy) have to encourage the resources on the demand side, as Demand Response, to participate together with the generation to the markets, and indicates that the system operators have to consider the Demand Response providers, included the aggregators, in non- discriminatory mode, on the basis of their technical possibilities, satisfying the system constraints P. Bertoldi, P. Zancanella, B. Boza-Kiss, Demand Response status in EU Member States, JRC Science for Policy report, 2016 © Copyright Gianfranco Chicco, 2012-2020  Demand Response Development in Europe § Year 2017 http://www.smarten.eu/wp-content/uploads/2017/04/SEDC-Explicit-Demand-Response-in-Europe-Mapping-the-Markets-2017.pdf © Copyright Gianfranco Chicco, 2017-2020  DR Access to Markets in Europe § Year 2017 http://www.smarten.eu/wp-content/uploads/2017/04/SEDC-Explicit-Demand-Response-in-Europe-Mapping-the-Markets-2017.pdf © Copyright Gianfranco Chicco, 2017-2020  Demand Response Programmes in Europe § In Europe the development of DR programmes is still relatively limited (the active ones are prevailingly load control and interruptability programmes for industrial users), also because of: § High implementation costs § Reduced information on the possible benefits from their use § Trend of the users to avoid changes in their lifestyle (with “storage” of resources of various types to reduce the possible discomfort) § Limited amount of the possible benefits in some cases (e.g., reductions in the electricity prices for households) § Limited dedicated regulatory rules © Copyright Gianfranco Chicco, 2017-2020  Demand Response in Europe n Conclusions from the Smart Energy Demand Coalition Report (2017): q The regulatory framework in Europe for Demand Response is progressing, but further regulatory improvements are needed q Restricted consumer access to Demand Response service providers remains a barrier to the effective functioning of the market q Significant progress has been made in opening balancing markets to demand-side resources q The wholesale market must be further opened to demand-side resources q Local System Services are not yet commercially tradeable in European countries http://www.smarten.eu/wp-content/uploads/2017/04/SEDC-Explicit-Demand-Response-in-Europe-Mapping-the-Markets-2017.pdf © Copyright Gianfranco Chicco, 2012-2020  DR BASELINE AND DR PERFORMANCE  The importance of the baseline n The baseline is the reference pattern associated to a customer or a customer group: q The true baseline is the pattern that a customer would have followed in the absence of a DR action q The predicted baseline is the pattern estimated by the utility company the customer (or customer group) would have followed in the absence of a DR event (i.e., it is the prediction of the true baseline) n Methods proposed to determine the predicted baseline: q HighXofY q MidXofY q exponential moving average q regression baseline n The baseline may also depend on temperature and weather ENERNOC White Paper, The Demand Response Baseline, 2009 T. K. Wijaya, M. Vasirani and K. Aberer, "When Bias Matters: An Economic Assessment of Demand Response Baselines for Residential Customers," IEEE Transactions on Smart Grid, vol. 5, no. 4, pp. 1755-1763, July 2014  The importance of the baseline n The performance of the DR action is assessed by calculating the difference between the actual pattern that followed the DR action and the corresponding “business as usual” baseline n The baseline and the DR outcomes have to be: q accurate, in order to reflect the real curtailments due to DR changes q defined with integrity, i.e., not containing irregular data and able to discover the presence of irregular responses to avoid customers “gaming” the system q simple, enabling straightforward calculation and interpretation q aligned with the DR goals, to avoid inadvertently penalising DR efforts n Balancing the above aspects is highly challenging ENERNOC White Paper, The Demand Response Baseline, 2009  Adjusted baseline n The adjusted baseline is the adaptation of an initial baseline to the actual load pattern occurring before the start of the DR action n Specific rules have to be followed to determine the adjusted baseline ENERNOC White Paper, The Demand Response Baseline, 2009  Determination of the adjusted baseline n Example of calculation of the adjusted baseline for a given time interval t: (e.g., 5-min time windows) The initial baseline bt is calculated as the average demand among the X=5 highest energy usage days out of the prior Y=10 non-event days bt = (ctd1+ctd2+ctd3+ctd4+ctd5) * 1/5 The adjustment factor at is the difference between the observed demand and estimated baseline for a calibration period starting two hours before the event notification, with a minimum adjustment of 0 at = max{[(ct-1–bt-1)+(ct-2–bt-2)]*1/2, 0} ENERNOC White Paper, The Demand Response Baseline, 2009  Determination of the performance variables n Considering each interval i over a DR event period beginning at time 0 and ending at time e: The total performance p is the integrated difference between the sum of the baseline bi and the adjustment factor a (constant during the event period) less the consumption ci The capacity-setting performance pavg is the average performance during all intervals of the DR event where the program rules stipulate that performance is mandatory pavg = p / e ENERNOC White Paper, The Demand Response Baseline, 2009  Strategic behaviour n Strategic behaviour to artificially create favourable conditions in the determination of the baseline, leading to economic advantages in the determination of the reward after a DR event, has to be avoided n The possibility of conducting strategic activities impacting on the baseline depends on the baseline adjustment approach and increases when the DR event is announced well in advance to the participants n For example, considering ad additive adjustment the user could increase the load a few hours before the DR event, in order to have a higher baseline on the basis of which the demand reduction will be calculated n Pre-cooling of the load is one possibility of increasing the load before the DR event, then cutting the supply to the cooling load during the DR event n Appropriate rules have to be established to identify and mitigate the strategic behaviour KEMA, Inc., PJM Empirical Analysis of Demand Response Baseline Methods, White Paper, 2011