218964 chromalin 20160216155100the business model of utilities, new business models can shape future...

FEBRUARY 2016: Issue 104

A QUARTERLY JOURNAL FOR DEBATING ENERGY ISSUES AND POLICIES

forumElectricity is in ferment – an unusualstate for an industry which hastraditionally enjoyed the security of long-term assets, steadily growing demand,and stable revenues (which has, in short,long been the position of a typical utility).These secure foundations are nowcoming into question as the industryfaces major technological, economic,and institutional change. Perhapsmost visible are the developments inelectricity generation – the growingpenetration of intermittent renewableplants, driven both by technologicaladvances and by the policy commitmentto decarbonization. But significant shiftsare also taking place elsewhere in thesystem with the rapid development ofinformation and control technology,which is opening the way for newapproaches to system management andmore flexible demand. It is likely that weare only seeing the beginnings of thesechanges – they raise wider questionsabout the very nature of the industry’sproduct and its relationship with itscustomers.

The technological developments havebeen accompanied by major policyand economic changes – fallingelectricity demand, greater use ofon-site generation leading to lowernetwork income, governments ratherthan markets driving investment in bothrenewable and fossil generation, and

so on. The institutional frameworkssurrounding the industry are strugglingto keep up. For two decades or soafter 1990, governments across theworld focused on liberalization and theextension of market forces; now thereis a new emphasis on decarbonization,but governments have not yet workedout whether decarbonization andliberalization can go hand in handor whether there is a fundamentalconflict. Markets have also been slowto adapt to the new era – the industryhas traditionally relied on marginalcost (kWh) pricing, although a largeproportion of its costs have alwaysbeen fixed (and some eminenteconomists like Ronald Coase haveargued against the over-emphasison marginal cost). With a growingpenetration of zero marginal costplants, the marginal cost approachlooks increasingly outdated, whetherat wholesale or retail level. Regulationtoo needs to respond to the changes –including the increasing decentralizationof the system – under way. Newcoordination and control methodsmay be required to manage the rapidgrowth of intermittent generation,particularly wind. Indeed the wholebasis of the industry’s workings iscoming into question: what ultimatelyare its products? How should it pricethem? What business models shouldthe industry be developing? What are

OXFORD ENERGY FORUM

CONTENTS

Transformation of the ElectricitySector: Technology, Policy andBusiness Models

Future business models for power markets:what can we learn from the ‘sharing economy’?Rolando Fuentes 5

How the ‘Big Beyond’ will change businessmodels of utilitiesChristoph Burger and Jens Weinmann 8

Renewable integration and the changingrequirement of grid management in thetwenty-first centuryRahmat Poudineh 11

Market integration with energy-only marketsand renewables: lessons of experience fromAustraliaRabindra Nepal and Tooraj Jamasb 14

Reforming European electricity markets forrenewablesRichard Green and Goran Strbac 17

Supporting renewable generation in the UKDavid M. Newbery 19

Centralized or decentralized? Remove firstthe regulatory barriersIgnacio Pérez-Arriaga and Scott Burger 22

Wind power and electricity supply securityin ColombiaDavid Harbord, David Robinson, and Iván M. Giraldo 26

Decarbonization through electrification – theimportance of energy taxation being in linewith long-term energy policyGraham Weale 29

View from the USA: rapid transformation ofpower sector calls for new tariffsFereidoon P. Sioshansi 33

A comparison of US and EU electricity prices:the relevance of the government wedgeDavid Robinson 35

Assessment of demand response in ShanghaiMalcolm Keay, Xin Li, and David Robinson 40

Power system reform to enable large-scalewind utilization in ChinaZhang Xiliang and Xiong Weiming 43

INTRODUCTIONTO

THISISSUE

its resources and how do storage anddemand response fit in?

These fundamental questions underliemany of the articles in this issue ofthe Oxford Energy Forum. One themeis the need to look at basic businessmodels. Rolando Fuentes considers theimpact of distributed energy resources(DER) such as solar panels, batteries,smart meters, and appliances, whichare together acting as disruptivetechnologies. Using two examplesof successful business models fromthe ‘sharing economy’ – UBER (taxi)and Airbnb (room letting) – the authoridentifies three big issues for thepower sector: the reallocation of risk,the paradoxical role of price signalsin situations of spare capacity, andthe dynamics between regulation andbusiness models. He argues that,contrary to the current situation inwhich electricity regulation dictatesthe business model of utilities, newbusiness models can shape futureregulation; by re-allocating risks, anew role can be created for incumbentutilities. Fuentes also proposes thatelectricity markets need to be re-designed to take into account themultidimensional nature of electricityand distinguish between separateservices such as reliability (MW),energy (MWh), system savings (NWh),and environmental benefits (carbondioxide emissions reductions).

Christoph Burger and Jens Weinmannalso look at the implications of thegrowth in renewable generation forbusiness models, particularly inGermany. Germany is in many waysaffected by the issue more acutelythan other countries, both becauseit has seen such a rapid increase ingeneration from wind and solar PV,and because in Germany little of thisnew generation is owned by the mainutilities themselves. Those utilities aretherefore facing a major erosion oftheir revenues and need to come upwith imaginative responses – a difficult

task for companies used to moreconservative ways of thinking. However,they are exploring new approaches,like outsourcing innovation andpromoting co-investment and venturecapital funds. But the long-termchallenges posed by IT may be evenmore fundamental – challenges suchas the Internet of Things, the sharingeconomy, and the use of blockchainsto squeeze out intermediaries.

Integration of the new sources andsystem coordination are the focus of anumber of articles. Rahmat Poudinehdiscusses the new requirements forgrid management that result fromthe integration of renewable energyresources. In recent years supply sidefluctuations, mainly related to wind andsolar power, have presented a newform of uncertainty to the power systemand added to the complexity of gridmanagement. Power systems with alot of renewables tend to experienceshort and steep ramps, such that thesystem operators need to bring on orshut down power plants more oftenthan in the past, and within very shortperiods of time. He contends thatas intermittency increases and theneed for flexibility becomes critical,the system will require new models ofgrid management both at operationaland institutional levels. In additionto traditional resource adequacymetrics we need new methodologiesto quantify and assess the technicallyavailable operational flexibility of thepower system. In addition, he argues,we need to ensure that electricitymarkets and regulations not onlyprovide incentives for investmentin flexible resources but also makeefficient use of available resources.

Rabindra Nepal and Tooraj Jamasbdiscuss the experience of marketintegration, renewable energy, andnetwork regulation that has beenbuilt up by the Australian NationalElectricity Market (NEM). They providean overview of the structure and

organization of the NEM and pointout that, more than 14 years since theestablishment of the NEM, wholesaleelectricity price differences persistacross the Australian states. The mainreason for this is the presence ofnetwork constraints across the regionalinterconnectors, which impedesthe wholesale market integrationprocess in the NEM. In addition,the authors argue that the lack ofadequate interconnection prevents thedevelopment of new wind power in theresource-rich regions and leads to thecurtailment of the existing capacity.The costs in terms of lost revenueto wind power plants as a result ofcurtailment are large and hampermeeting the national renewableenergy targets. Therefore, the authorscontend that the development ofrenewable energy and electricity marketintegration can be complementarypolicy objectives in the NEM as one ofthe largest interconnected energy-onlymarkets in the world.

Richard Green and Goran Strbacexplore ways of adapting Europeanelectricity markets for renewables.The authors argue that nationalrenewable energy policies in the EUregion are still uncoordinated, and theconcentration of renewables in morefavourable locations would bringimportant savings. Additionally, theyargue that the existing linkagesbetween day-ahead markets need tobe supplemented with greatercoordination on longer and shortertimescales. Short-term balancingcosts could be reduced if systemoperators had access to balancingservices and reserves in neighbouringsystems. Cross-border trading oflong-term contracts for energy wouldbe facilitated by the use of FinancialTransmission Rights (FTRs), which donot have the restrictions of capacitycontracts. They also suggest that inorder to facilitate balancing betweencountries, we may have to integrate

TRANSFORMATION OF THE ELECTRICITY SECTOR

2 OXFORD ENERGY FORUM

energy and reserve markets withineach country – something that theAmerican experience shows can bedone. However, the politicalchallenges involved in creating a goodmarket design should not be under-estimated.

Regulation and government policy alsoneed to be addressed. David Newberyexamines the logic, drawbacks,and possible resolutions of the EUCommission’s proposed shift fromfeed-in tariffs (FiTs) to Premium FiTs(PFiTs), for future UK renewablessupport. The author reviews thehistory of UK renewables support andargues that an important advantageof quantity-based instruments likeRenewable Obligation Certificates(ROCs) is that their cost is controllable.However, ROCs are exposed tovolatile market prices and abundantrenewables deployment can depresstheir prices. In contrast, a fixed FiTprovides greater insurance but at thesame time risks excessive demandand budgetary cost, if no cap is set onthe total quantity procured. From theauthor’s point of view, the question ishow to combine the advantages of aPFiT with the risk-reducing propertiesof a FiT. Newbery argues that capacitycontracts with suitable Power PurchaseAgreements (PPAs) and contracts forancillary and balancing services (withall contracts reflecting efficient marketvalue) are the logical solution andshould support the delivery of low-carbon electricity at least cost.

Ignacio Pérez-Arriaga and Scott Burgeraddress the regulatory challengesrelated to the growing penetrationof distributed energy resources(DER) such as: gas-fired distributedgeneration, solar PV, small- andmedium-sized wind farms, electricvehicles, energy storage, and demand-side management. The fundamentalquestion is what role should be playedby DER, and what by conventionalcentralized energy resources. They

argue that current regulations arewoefully inadequate to meet theincoming challenges and that whatis needed is an in-depth review andcorresponding modification to createan economically neutral playing field,enabling centralized and decentralizedresources to compete and collaborateefficiently, while recognizing that theyperform under very different conditions.This regulatory review should applythe fundamentals of microeconomicsto power systems in order to do thefollowing: reconsider the definitionof ‘essential electricity services’;examine how to compute prices so thateconomic efficiency is maximized; andunderstand what, if any, is the valuethat aggregation of DERs and anyassociated business models bring tothe power system.

David Harbord, David Robinson, andIvan Giraldo illustrate how the energyregulatory regime in Colombia acts asa barrier to investment in wind power.The Colombian electricity systemrelies very heavily on hydro power.However, during El Niño (dry) weatherperiods, hydro-based generationfalls significantly. Consequently, theregulatory system has been designedto finance the building of generationcapacity that can substitute for hydroduring dry periods. The regulatororganizes auctions to select theplants that will be awarded long-termcontracts to provide backup, and todetermine the price for ‘firm energy’.A critical issue has to do with howmuch firm energy is attributed todifferent technologies in the auctions.The authors argue that the currentmethodology underestimates theamount of firm energy available fromwind power because it does notadequately reflect the evidence thatwind generation is greater when itis most valuable, specifically duringEl Niño periods. Since secure long-term revenues are directly related tothe amount of firm energy sold in the

auction, the regulations therefore act asa barrier to investment in wind power.

Electricity pricing and taxation arefacing major new challenges. GrahamWeale considers a significant policyconsistency. On the one hand,governments see electricity as themain vector of decarbonization, via agreatly enhanced role for decarbonizedelectricity within energy supply. On theother hand, current taxation policy ishindering the process of electrificationby loading more tax and surchargeson to electricity than on other energysources – in Germany, taxes andsurcharges on electricity for electricvehicles are actually higher than taxeson vehicle fuels. One result is that inmany countries, electricity’s marketshare is actually starting to decline.Weale considers various options forbringing energy taxation into line withoverall energy and climate policy.Covering the costs of decarbonizationfrom general taxation seems like alogical solution but it raises manypolitical difficulties – the importantthing is to recognize and address theproblem.

Fereidoon Sioshansi looks at theimplications of the changes under wayin the power sector for the structure ofelectricity tariffs, focusing on the USA.Most consumers there pay on avolumetric (kWh) basis. This meansthat utilities lose income whenconsumers install on-site generation,such as rooftop solar PV, and as aresult buy less from the grid. To makeup for the loss of income, utilities maybe forced to raise their tariffs – thusfurther encouraging investment inenergy efficiency and on-sitegeneration and leading to theso-called utility ‘death spiral’. Theprocess leads to a shift of costs fromcustomers with solar generation tocustomers without, raising significantequity issues and pitting the two setsof customers against each other.Sioshansi considers various possible

INTRODUCTIONTO

THISISSUE

FEBRUARY 2016: ISSUE 104

3OXFORD ENERGY FORUM

solutions, such as three-part tariffs,increased fixed charges, time-of-userates, and demand subscriptions, butconcludes that ‘the battle is justbeginning’.

David Robinson analyses the reasonswhy final electricity prices in the EUhave risen so much faster than inthe USA since 2008. This cannot beexplained by different trends in costsrelated to wholesale generation,transmission and distributionnetworks, and retail services. Rather,it reflects the increase in the EU of a‘government wedge’ in final electricityprices. This wedge includes taxesand levies related to the financing ofdecarbonization, and also supportfor domestic fuels (such as coal),subsidies for specific consumergroups, and other policy objectives.In short, final electricity prices in manyEU countries no longer reflect thecosts of supply, but rather the costs ofmeeting a variety of policy objectives.This has distributional consequencesand hurts the competitiveness of manyindustrial and commercial consumers.Furthermore, it introduces economicdistortions – which include an incentivefor consumers to favour fossil fuelsover electricity. He makes a numberof policy recommendations, includingthe introduction of carbon taxes and

shifting some of the policy costs to thegovernment budget.

Finally, two articles look at experiencein China, which is facing challengessimilar to those elsewhere in the worldin relation to integrating the newresources, and is responding via amove to greater reliance on economicsignals and market mechanisms.Malcolm Keay, Xin Li, and DavidRobinson report on a studyundertaken by the OIES, along with theEnvironmental Change Institute ofOxford University, on the potential fordemand response in Shanghai. Whileinterest in demand response isgrowing worldwide, examples fromoutside the OECD are often lessfamiliar. Furthermore, the Chineseelectricity system presents manyspecial characteristics, in particularthe widespread use of a form ofadministrative demand control tobalance the system. Nonetheless,China is moving towards a greaterreliance on economic instruments inelectricity, as in other sectors, and theShanghai study aimed to establishwhether demand response could havea role there, as elsewhere. It showedthat there is indeed potential for thisapproach, which could have botheconomic and environmentaladvantages, and the issues involved

are now being examined in more detail.

Zhang Xiliang and Xiong Weiminganalyse the significant curtailment ofwind power in China. Although Chinahas become the largest wind powermarket in the world, in the first half of2015 over 15 per cent of technicallyavailable wind power could not beconnected to the grid and had toshut down. The reasons includeuncoordinated planning between thegrid and the wind generators, inflexibledispatch, and the lack of suitablepolicy incentives for stakeholders,notably the grid operator and the coalpower generators. The governmenthas introduced a number of policyreforms, but the curtailment problemremains serious. In May 2015, the StateCouncil released Document 9, whichcould be the basis for deeper reformsof the power sector. It emphasizes therole of market mechanisms, especiallyin generation and retail segments,and also stresses the importance ofintegrating renewable energy in orderto facilitate decarbonization, improveair quality, and mitigate climate change.The authors conclude that the bestway to solve the problem of curtailmentof wind power is to introduce marketmechanisms, such as a spot market,as well as policy incentives for keystakeholders.

INTRODUCTIONTO

THISISSUE



Future business models for power markets: what can we learn from the‘sharing economy’?Rolando Fuentes

Innovations in distributed energyresources (DER) − like solar panelsand batteries, information technologies,smart meters and appliances − havethe potential to rock the fundamentalsof the electricity sector. It now seemslikely that a significant proportion offuture end-use electricity consumptioncould be supplied and managed byrelatively small-scale, distributedresources, opening up attractivealternatives to traditional utilities forcustomers. This shift could reducereliance on the central grid, whichcould ultimately change the wayelectricity is purchased, transported,and consumed. Boundaries betweentransmission–distribution anddistribution–commercialization andgeneration may become blurred, asthese activities would occur in thesame place: the household. However,the system would become morecomplex, since households aregeographically dispersed.

‘…A SIGNIFICANT PROPORTION

OF FUTURE END-USE ELECTRICITY

CONSUMPTION COULD BE SUPPLIED AND

MANAGED BY RELATIVELY SMALL-SCALE,

DISTRIBUTED RESOURCES…’

The arrival of these innovations isleading utilities around the world tore-evaluate their business models, andregulators are considering electricitymarket reforms. However, trying to fixthe system with marginal or incrementalamendments will not be sufficient tocope with all these changes. In thisarticle we explore experiences gatheredfrom other industries that have facedtechnological disruptions recently,together with lessons they have learnedwhich might guide future businessmodels in the power sector.

Value proposition

There are at least three reasons torevisit business models.

1 These technological disruptions makeit more evident that electricity is amulti-dimensional commodity, theattributes of which at times may bedifficult to price. For example, boundup with the energy service itself are:efforts to reduce emissions, reliabilityservices, and risks. Also, the non-consumption of power has a value tothe system. A key question for futurebusiness models then is how tomonetize these values.

2 This eruption of innovative newtechnologies is taking place in asector that has sunk costs and whereinvestments are already in place. Hencethe power sector cannot start afresh.

3 There is a place for both utilities andnew entrants in the industry, albeitwith different roles.

The prevailing business model forelectricity utilities is a cost-plus structurein which the utilities pass their costs,plus a return on their capital investment,to customers at a variable rate (USD/kWh). The objective is to operate in acost-minimization fashion and the modelsustains itself with further capitalinvestments, sales growth, andsustainable prices. This has led to thedevelopment of a business model whereadding new capacity is the bread andbutter of utilities’ revenues. But shouldwe still impose this framework on utilitiesin the future, given massive investmentrequirements and lower sales?

There is a legacy in the power sectorand the impact of the adoption of thesetechnologies could go either way, witha positive or a negative impact.Uncoordinated introduction of DER

could increase system risks andtransfer costs to other customers, in theabsence of an organized market.However, so long as these newertechnologies operate in a coordinatedway with the other power sectorresources, they can provide value tocustomers and to the overall system.Also, the value delivered by distributedtechnologies together would be greaterthan the sum of the values delivered byindividual components.

These facts lead us to posit different rolesfor incumbent utilities and for newentrants. A utility offers not only energy toits customers but also spare generationcapacity, ramping flexibility, operatingreserves, spare distribution capacity,and ancillary services. But customers donot value all these items in themselves,since they don’t see them or think aboutthem. Rather they value the servicesthat depend on them – for example,reliability – from their power provider.

Airbnb and UBER

It has been suggested that solarphotovoltaic (PV) panels, batteries, andsmart grids can essentially transform thepower markets into a series of nestedmarkets, connected through differentplatforms as if they were multiple-sidedmarkets. According to an IDEI (InstitutD’Economie Industrielle) working paper of2004 – ‘Two-Sided Markets: An Overview’by Jean-Charles Rochet and Jean Tirole– a multiple-sided market is a meetingplace for a number of agents that interactthrough an intermediary or platform.Markets of this sort, such as credit cardcompanies and stock exchanges, havebecome prevalent in today’s economy.Google Search, Amazon, Facebook,UBER, and Airbnb are just a few of themore prominent examples. In these types



of markets, according to the EPRG(Energy Policy Research Group) workingpaper ‘Platform Markets and EnergyServices’ of 2013 by Claire Weiller andMichael Pollitt, an intermediary capturesthe value of the interaction between usergroups, and network externalities maylead to one of them being charged a noncost-reflective price.

‘THE VALUE PROPOSITION OF THE

SHARING ECONOMY BRINGS TOGETHER

OPERATORS WITH UNDERUSED ASSETS…’

The UBER (taxi) and Airbnb (roomletting) business models – twosuccessful companies of the ‘sharingeconomy’, as it has come to beknown – may help us understand thefuture business models of the powergeneration sector. We can draw threelessons from the sharing economyexperience. The first is that risk isre-allocated across the market, thesecond is the paradox between sparecapacity and price signal, and thethird relates to the dynamics betweenregulations and business models.

The value proposition of the sharingeconomy brings together operators withunderused assets and others that maywish to hire or rent, in a timely manner,with low transaction costs since theirinformation is more transparent dueto the use of an internet platform. Thebusinesses of the sharing economy setout to optimize the use of resources bymaking more frequent use of excesscapacity in goods and services.According to an Economist article (‘TheRise of the Sharing Economy’) in March2013, other terms that have been coinedin the media for this sector are the‘collaborative economy’, the ‘asset-lightlifestyle’, or the ‘access economy’.

Risk-reallocation

One of the key elements of the UBERor Airbnb business models is the needto deal with the uncertainty buyers and

sellers have in sharing and trading withpeople they have never met. Thus, as aside effect, insurance companies havegained from this model as more peopledemand their service. Borrowing this idea,an incumbent utility could act as theinsurer of the rest of the power sector. Asmore renewable energy is installed andoperated, this would mean two things:a growing share of generation would beat zero marginal cost, and utilities couldend up with unused capacity for longperiods of time. Paradoxically, that wouldmake traditional capacity more importantsystem-wise, as the incumbent would bethe operator of last resort to keep thelights on. Consequently, the benefit ofmaintaining the grid’s operability with thecostliest energy source would outweighthe cost of the absence of electricity.

If utilities are to use their infrastructureas insurance, they will need to changethe way they charge customers. Ahealth insurance company’s businessmodel, for example, is based onhealthy people financing treatmentsfor ill people. The way forward for theutility could be to charge a fixed price,like an option value, for back-up. Thiswould compensate for the long periodsof zero marginal cost generation thatcould undermine the energy servicetheir installed capacity provides, but notthe reliability component provided bythe same capacity. Utilities could offermemberships, or stream services like theinternet television network Netflix, wherecustomers pay a fixed amount. Or energyproviders could offer contracts of servicefor electricity ‘on demand’, which wouldbe more expensive, or ‘as available’,based on solar or other intermittentrenewable generation (according to thechapter ‘Electricity markets and pricingfor the distributed generation era’ byMalcolm Keay, John Rhys, and DavidRobinson in Fereidoon P. Sioshansi’s2014 book Distributed Generation andits Implications for the Utility Industry).

This option would detach power pricesfrom the time variable. The alternative

would run contrary to the recenttendency to have more real-timedecisions with smart metering, as itwould decrease the number oftransactions and increase the timeperiod during which they take place.Counterintuitively, this option wouldincrease the time lags of transactions bycharging fixed prices for coverage inlonger periods. However, fixed priceshave been met with significant resistancefrom energy efficiency and DERadvocates, because customers wouldno longer have the incentive to adopt thenew technologies and thus save money.

There are two caveats. First, as is oftenthe case, the starting point matters.For example, people are willing to payfor health insurance to increase theircoverage, whereas customers of thepower sector start with near 100 percent coverage already. In completeenergy autonomy they would need to getused to paying for a service that they areaccustomed to receiving and taking forgranted, alternatively, they would need todemonstrate their willingness to acceptoccasional service disruptions. Second,contracts for stream services like Netflixare feasible since there is no rivalry inconsumption in their stream service.Electricity is a rival good in consumption,but it can still be argued that not all of theutility’s customers will need back-up atthe same time, just as not all people claimon their health insurance or cash-in theirbank accounts at the same time.

Spare versus scarce

There is a paradox in the sharingeconomy model. The traded asset isusually idle capacity − for example,Airbnb’s asset is unused rooms inpeople’s houses − whereas economictheory focuses on how prices sendscarcity signals. The question is: whatare the spare, and what are the scarce,assets in the new electricity sectorstructure? It may be that the abundantasset will be PV solar panels. One of theconcerns is that PV ends up completely



flooding the electric power system withuncontrolled amounts of zero variablecost energy. The scarce resource ismore difficult to work out. However,bearing in mind the nature of the newtechnology, the distribution sector, themultidimensional characteristics ofproducts and services in the powermarkets, and the network effects, weforesee the following:

Average cost per kW will be lower. Arecent multidisciplinary MIT study(‘The Future of Solar Energy’, 2015)shows that a large penetration ofsolar PV displaces the plants with themost expensive variable costs andincreases the cycling requirements ofthermal plants.

The role of energy aggregator will bemore widely seen. If generation isintermittent, dispersed, and uncertain,what business ideas can amelioratethese undesirable characteristics?The answer lies in the creation of arelatively new player in the market:the aggregator. The aggregator’sfunction would be to coordinate theloads from dispersed sources, thussmoothing out the underlyingintermittency. These new players arealready emerging in California.

Indirect services will be central to thenew electricity markets. Electricityhas multidimensional attributes. Themost straightforward is the quantityof energy delivered at specific timesand locations. But we can alsoconsider the purpose for which theelectricity is used. For example:charging a battery, running a fridge,watching TV, end-use, cooling, andheating. And we should consider itsreliability − the probability that supplywould be available. Firms can sellreliability (MW), energy (MWh),

system savings (NWh), orenvironmental benefits (carbondioxide emissions reductions).

Managing demand profile is key for thenew marketplace: the distribution. Oneof the main determinants of distributionoperations is demand profile. How canbusiness solutions, to some extent,help to manipulate demand profilesso that non-consumption anddemand response become valuable?

A key issue would be how to pricethese services, as most of them areexternalities – positive and negative.There are a couple of alternatives:

Regulation that imposes values,products, and transaction guidelines.Such regulation would establish therules of the markets, while themarkets themselves would thendecide the efficient level of provision.

A Coasian approach wheregovernments assign property rights,for example over the reliability aspectof the system, and permit actors totrade between themselves.

Regulatory dynamics

Utilities’ business models have largelybeen dictated by their regulatory model.It is true that the rate-of-return regulationin some countries has made the utilitybusiness model an infrastructure one.But in other regulated industries wheretechnological innovation has changed thelandscape, regulation has proved tohave been lagging behind the impetusof the industry. One of the most notablecases where regulation has been forcedto adapt to these technologies is in the taxiservices industry, with the advent of UBER.

In such a case, the dominant strategyfor a new entrant – UBER, or Airbnb,for example – is to quickly grab marketshare to lock in their platform. This wouldforce regulators to accommodate thesenew business practices as the standardregulation. For example, regulators havetaken a hands-off approach with regardto risks for new taxi services and have

chosen to include insurance clausesas part of their regulation. This wasadapted from UBER’s own businessmodel. Other jurisdictions facing similarproblems are likely to follow what thefirst mover regulators have done. Thiswill then become the norm. Having ahomogenous regulation framework helpsnew companies expand more rapidlyacross different cities or countries.A key element for the electricitygenerating industry is this chickenand egg question: will penetration ofdistributed energy resources occur inan orderly manner, allowing regulatorsto accommodate them? Or will newentrants and new business models in theend pre-empt future regulation?

Conclusion

Business models in the electricitysector have been under scrutiny. Inthis article we try to draw parallelsfrom industries that partially resemblethe power sector and that have facedrecent technological disruptions.

The main points we can draw fromthose are:

1 technical disruptions will forceregulations and business models toadapt;

2 business models can shape futureregulation, as opposed to the positionat present where regulation dictatesthe business models for utilities;

3 re-allocation of risks opens up theprospects of a new role for theincumbent utility; and

4 electricity has multi-dimensionalattributes for which markets need tobe designed.

This article is based on a forthcomingKAPSARC publication

‘THE AGGREGATOR’S FUNCTION WOULD

BE TO COORDINATE THE LOADS FROM

DISPERSED SOURCES, THUS SMOOTHING

OUT THE UNDERLYING INTERMITTENCY.’

‘WILL PENETRATION OF DISTRIBUTED

ENERGY RESOURCES OCCUR IN

AN ORDERLY MANNER ALLOWING

REGULATORS TO ACCOMMODATE THEM?’



Challenges of energy incumbents

According to figures from the Germanenvironmental ministry, Germanyincreased its intake of renewableenergies from 6 per cent in 2000 to 32.5per cent in the first half of 2015. Thecountry’s incentive system, based onfixed feed-in tariffs, motivated privateindividuals such as homeowners,farmers, and energy associations toparticipate in the market. In 2014,around half of all renewable energyinstallations were owned by citizens,rather than corporate entities, and morethan 1.5 million producers of electricityparticipated in the market.

‘IN 2014, AROUND HALF OF ALL

RENEWABLE ENERGY INSTALLATIONS

WERE OWNED BY CITIZENS, RATHER

THAN CORPORATE ENTITIES…’

In contrast, the ‘Big 4’ energyincumbents operating in Germany– EnBW, E.ON, RWE, and Vattenfall– own less than 10 per cent of allrenewables installations. One of thereasons why incumbents have notparticipated in the rush for renewableenergies may be that their businessmodel has traditionally been directedtowards large-scale projects, ratherthan smaller, decentralized installations,given that the proportion of transactioncosts involved in small-scaleinvestments is typically higher than inlarger projects.

Since the marginal costs of sun andwind power are practically zero, thepriority feed in of renewable energieshas induced a downward pressureon prices in the German wholesalemarket. Average peak load prices havedecreased from around 60 EUR/kWh in2011 to less than 35 EUR/kWh in 2014,according to the Fraunhofer Institute

for Solar Energy Systems (ISE). E.ON,the largest German electricity producer,posted a loss of EUR5.7 billion forthe first nine months of 2015, whilecompetitor RWE’s operating resultdecreased by 9 per cent. The erosionof revenues has led to a strategic shiftbeing undertaken by E.ON and RWE.Both utilities intend to split their assetsinto two separate and independententities, with the intention ofreorganizing their activities and creatingcompanies that are able to fit better intothe changing marketplace.

Counter-strategies and new forms ofinnovation

Many European electric utilities thatare exposed to similar developmentshave realized that their business modelhas to change if they want to survivein the competitive environment. Theyhave to become more innovativeand explore new business models tocompensate for the loss of revenuesin their traditional business units. Inthe past, research and developmentwithin utilities predominantly focusedon incremental innovation, on thegradual improvement of processes andoperations, rather than on disruptivebusiness ideas. Utilities with a retailand distribution business have aparticular advantage in the testing ofnew business models, because theyare already physically present onthe premises of their customers, viametering technology.

‘MANY EUROPEAN ELECTRIC UTILITIES …

HAVE REALIZED THAT THEIR BUSINESS

MODEL HAS TO CHANGE IF THEY WANT

TO SURVIVE…’

However, utilities struggle to findpeople within their workforce that are

capable of developing genuinely new,customer-centric, digital businessmodels. Hence, they follow a strategythat has proven successful in manyother industries, as diverse aspharmaceuticals, logistics, or finance,over the last couple of years. Theyoutsource innovation to young teamsof entrepreneurs and the founders ofnew companies. The most commoninstruments of that process are venturecapital (VC) funds, incubators, andaccelerators. VC funds typically providefinancial resources to a portfolio ofsmall companies, but they do notinterfere with the day-to-day operationsof their ventures. By contrast, corporateincubators provide an environment (butnot necessarily a physical location)where new ideas from inside or outsidethe company are nurtured, protectedfrom the company’s other keyperformance indicators and standardcorporate culture. Accelerators, thethird common type of these new formsof innovation, are programmes of twoto six months duration in which a teamof mentors and coaches guides thefounders of new companies throughthe process of developing a businessplan and of finding investors tocommercialize their ideas.

Hybrid forms of business model arealso possible: E.ON, Germany’slargest utility, pursues a strategyof co-investments where (limited)involvement and early bondingbetween a business unit (which takesover some type of ‘ownership’) and theventure is desired.

As early as 2008, the Spanish energyutility Iberdrola began setting up a VCfund and through to 2015 most majorEuropean players followed – eitherestablishing their own organizationsor participating in collaborative efforts

How the ‘Big Beyond’ will change business models of utilitiesChristoph Burger and Jens Weinmann



with other companies. The diagrambelow shows a selection of initiativesundertaken by major European energyincumbents.

Three major challenges emerge withthis type of diverse innovation:

1 The multitude of ideas implies that alarge fraction of these start-ups andventures will not succeed in themarketplace. A venture capital firmtypically receives 2,000 businessplans, evaluates 200, and invests in20, of which roughly 2, or 10 per cent,outperform – a chance of 1:1,000 toidentify a successful business model.The ‘fail-fast’ attitude is an inherentcharacteristic of the start-upecosystem, but it requires afundamental change in mentality of thetop management of electricity utilities.

2 Integrating the ventures into existingorganizational structures is most likelyto succeed when done early. Forexample, German incumbent E.ONtries, very early in the process, toidentify business units that will take‘ownership’ of idea and start-up.

3 Will the new ventures yield sufficientrevenues soon enough to

compensate for losses in other units,thus ensuring corporate survival? TheUS company IBM is a prime exampleof a company that can reinvent itself,but many other incumbents fromother industries have disappeared.Polaroid, Kodak, and DEC are justsome examples.

Not all seems lost for the electricitysupply industry. While revenuesfrom generation decline, the grid-related business of utilities (electricitytransmission and distribution) securesa regulated (and capped) but stablestream of income. Utilities thatown parts of the grid are in a morecomfortable position than those thathave been forced to unbundle or tocreate separate legal entities.

In the longer term, however, thegrid-related services of utilities maybe threatened in the same way astheir generation business. Electricitycan now be generated at the point ofconsumption, not just in centralizedpower plants, while costs of renewableenergy installations have substantiallydecreased over the last couple ofyears. The European Union’s researchunit, the Joint Research Centre,

foresees electricity generation costs ofaround 0.02 EUR/kWh for photovoltaicinstallations that have already beenwritten off; this is cheaper than anyconventional generation technologyexcept large hydropower. At a retail priceof around 0.30 EUR/kWh for residentialconsumers in Germany, 0.28 EUR/kWhcould potentially be spent on batterystorage or other means to detachindividual customers from the grid.

The ‘Big Beyond’

Three overarching trends, which theauthors call the ‘Big Beyond’, may leadto an erosion of the market position ofincumbents, even in the regulated partsof their operations:

1 Frugal innovation and the Internet ofThings (IoT) – a low-cost alternativeto conventional solutions, especiallyin building efficiency;

2 The blockchain – a securetransaction technology for smartcontracts;

3 The sharing economy – creatingown networks, for example viafinancing but also operating.

Frugal innovations are innovationsthat are made at lower cost and lowercomplexity than standard solutions.Often they can be found in thecontext of developing economies,where many products used in theindustrialized world are too expensiveand over engineered. But even inwealthy countries, digitalizationcreates opportunities for founders andentrepreneurs to develop solutionsfor market segments that have beenuntapped, often in so-called legacymarkets. Retrofitting existing buildingstock and increasing the energyefficiency of houses is one of theareas where frugal innovation takesplace. Envio Systems, for example, isa Berlin-based start-up that intendsto revolutionize existing commercialbuildings. The company has developed

Timeline of exemplary initiatives of selected European utilities in thefield of ‘diverse’ innovationSource: based on C. Burger, S. Pandit, and J. Weinmann (2015). ESMT innovation index 2014 – electricity supplyindustry: The ‘big beyond’, ESMT Business Brief No. BB–15–01, (Figure 7).



a box called Cube that containsmultiple sensors and connects to adigital optimization platform. In contrastto, say, Google’s residential buildingefficiency device Nest, Cube is ableto detect whether there is anyone in aparticular room, via a carbon dioxidesensor, and can adjust the heatingor cooling activity accordingly. EnvioSystem’s solution costs a quarter ofsimilar installations offered by largecompetitors, such as SchneiderElectric, Siemens, or Honeywell. Thecompany estimates the market sizein North America and Europe at morethan 10 million commercial buildings.

If newcomers like Envio use theenhanced capabilities of devicesconnected via the Internet of Things,utilities operating in retail may loserevenues from an increasing numberof lucrative commercial customers.Going one step further, if costs forbatteries continue to decrease andstorage becomes affordable on a largerscale, commercial, industrial, and evenresidential customers may decide torely completely on self-production.

The blockchain is a secure transactiondatabase that is currently used forcryptocurrencies such as Bitcoin.Due to its configuration, it recordsall transactions ever executed in atransparent, decentralized internetprotocol. Its main differentiating driveris that it can replace all intermediaryinstitutions whose existence is primarilyjustified by their trustworthiness forall parties involved in the transaction.Banks belong to that group ofintermediaries, but so do energy utilitiesacting as retailers or traders.

This technology may be particularlyattractive in developing countries,where the poor often have no bankaccount and could potentially use theplatform as a means to pay their bills.For example, South African start-upBankymoon has introduced a systemthat allows users of smart meters topay via the cryptocurrency Bitcoin,thereby bypassing banks.

New venture Grid Singularity intendsto use the blockchain to introduce adecentralized energy data exchangeplatform. This would offer servicesenabling forecasting for grid balancingand would facilitate investment,the trade of green certificates, andeventually energy trade validation. GridSingularity’s founders expect immenseinfrastructure cost savings, compared

to standard technical solutions thatrequire a fully independent verticalintegration for each energy marketoperation. The platform may alsorender some of the current key entrybarriers to energy trade obsolete – suchas a need to have special accountsand deposits, or a certification withan intermediary financial institution.The business model of utilities maybe threatened if decentralized energyproducers enter into direct interactionwith their consumers. Any type ofphysical or monetary transaction thatinvolves a trustworthy intermediarycould be replaced by the blockchain.

The third element in the ‘Big Beyond’ iscrowdfunding, which is part of theso-called sharing economy: exclusiveownership of goods has become lessimportant for young people. The primeexample is car-sharing services, wherepeople rent cars for a limited amount oftime. The service – in this case,travelling from A to B – is the only factorthat counts. In the energy sector, founderspromote their business idea oncrowdfunding websites to collect money.Companies such as crowdEner.gy,econeers, Trillion Fund, or SunFunderestablish a parallel market that relies onthe financial commitment of manyindividuals, rather than on any bank orsingle investor. For example, the Germancrowdfunding company econeersrequires a minimum financial involvementof EUR250 over five years; investorsreceive dividends and are financiallyrewarded if the venture or project isprofitable. In 2015, econeers receivedEUR750,000 in solar crowdfunding.

Berlin-based start-up Sunride takesthe sharing economy one step furtherand implements software solutionsto optimize neighbourhood-levelphotovoltaic installations that arejointly owned and used by inhabitants.In Berlin, more than 120 localassociations have been founded to setup solar panels and generate electricityfor self-consumption and grid feed-in.

The ‘Big Beyond’ and exemplary companiesSource: based on C. Burger, S. Pandit, and J. Weinmann (2015). ESMT innovation index 2014 – electricity supplyindustry: The ‘big beyond’, ESMT Business Brief No. BB–15–01, (Figure 8).

‘[IF] STORAGE BECOMES AFFORDABLE ON

A LARGER SCALE, … CUSTOMERS MAY

DECIDE TO RELY COMPLETELY ON SELF-

PRODUCTION.’



The ‘Big Beyond’ not only questions thegrid-based business of utilities, bututilities in their very existence. Ifbuildings become largely energy-

autonomous via the IoT, all transactions

performed via blockchain, and finance

is undertaken via the sharing economy,

then utilities have to fear that they will

become the dinosaurs doomed for a

slow and painful decline of their

operations. By embracing the Big

Beyond, however, utilities might be able

to transform themselves into enablers

of a decentralized energy system. They

might even – rather than owning

large-scale assets, producing

electricity, and transmitting it to the

customer – build, own, or operate

decentralized energy generation

assets, steering them at the point of

consumption. It is only a matter of

technologies and mind set, since the

task – delivering energy to the

customer – would remain the same.

Renewable integration and the changing requirement of grid managementin the twenty-first centuryRahmat Poudineh

Background

Since the dawn of the electricityindustry, uncertainly has always beenan inherent characteristic of the system.The concept of uncertainty in the powersystem has evolved over time, butin the past it mainly implied demandfluctuation and components failure(generation and network). Traditionalsystems are equipped with a range ofcontrol mechanisms to manage thesekinds of variability and uncertainty, inorder to provide a reliable service. Insuch a paradigm, technologies aremature and the behaviour of loadis fairly predictable. In recent years,however, supply-side fluctuations havepresented a new form of uncertaintyand added to the complexity of gridmanagement. The source of supply-side variations is intermittency ofrenewable resources, such as wind andsolar, whose penetration is acceleratingas environmental regulations tightenaround the world. For example, thewind and solar PV installed capacity inthe EU region rose from 12.8 GW and0.125 GW in 2000, to 128.7 GW and 88GW respectively in 2014. This impliesthat solar PV constituted 10 per centand wind represented almost 15 percent of total installed capacity in the

EU in 2014. The rapid penetration ofrenewable resources imposes severaloperational and economic challengeson the electricity sector. The problemof the system operator is no longer toforecast load but net load, which is thedifference between the total load andthe supply from intermittent resources.This increases the need for systemflexibility; an ability that the powersystem requires in order to utilize itsresources and deal with variations innet load and generation outage overvarious time horizons.

‘THE RAPID PENETRATION OF

RENEWABLE RESOURCES IMPOSES

SEVERAL OPERATIONAL AND

ECONOMIC CHALLENGES ON THE

ELECTRICITY SECTOR.’

Integration of wind and solar resourcesnecessitates system flexibility in at leastthree ways.

1 The stochastic nature of variablegeneration leads to a widerconfidence interval of net loadforecast; it increases the need foradditional flexible resources. Althoughvariability declines with aggregationand geographic dispersion, no

forecast measure is perfect and errorwill exist even at an aggregated level.Moreover, even with perfectprediction, without curtailment, theavailable flexible resources may notbe sufficient to meet the variation inthe net load.

2 The penetration of intermittentresources can displace conventionalgeneration sources and this adverselyaffects the amount of online flexibleresources.

3 A lack of system flexibility can resultin more frequent occurrences ofnegative prices in wholesale markets;this imposes additional costs onrenewable energy levies (where theyexist) to cover the difference betweencontract and wholesale market prices(when guaranteed payments arelinked to the wholesale energy price).

Additionally, even if all mechanismsfor managing variability of net load areavailable, the current electricity marketsmay not be positioned to incentivizetheir efficient use. Thus, I contend thatas intermittency increases and theneed for flexibility becomes critical, thesystem requires new models of gridmanagement both at operational andinstitutional levels.

‘BY EMBRACING THE BIG BEYOND,

UTILITIES MIGHT BE ABLE TO TRANSFORM

THEMSELVES INTO ENABLERS OF A

DECENTRALIZED ENERGY SYSTEM.’



Intermittency: a driver of change in gridmanagement

In this section I briefly discuss some ofthe operational and institutional (marketand regulation) changes in the powersystem that result from the penetrationof intermittent resources.

Operational challenges

The operational challenges ofpenetration of variable generation intothe power system depend on variousfactors such as:

scale of intermittent resources,

correlation with demand,

flexibility of the power system.

In order to illustrate some of thesechallenges I use an example fromthe California Independent SystemOperator (CAISO) where the shareof renewables in the generation mixhas risen due to the ambitious goalsof the California Renewables PortfolioStandard (33 per cent of electricityfrom renewable sources by 2020 and50 per cent by 2030). These targets– along with environmental policiesand regulation regarding retiring ormothballing of the power plants thatuse coastal water – are expected tochange the energy landscape in favourof renewables in California.

‘TRADITIONAL RELIABILITY METRICS

NEED TO BE SUPPLEMENTED BY VARIABLE

GENERATION INTEGRATION STUDIES.’

The increase of intermittency hasalready started to manifest itself invarious operational challenges inCalifornia’s electricity sector. The CAISOpower system is now experiencing shortand steep ramps, such that the systemoperator needs to bring on or shut downpower plants more often and within veryshort periods of time, in order to meetthe variations in net load. Furthermore,during the hours of weak demand, theCAISO system runs the risk of over

generation. This usually happens whenthe system prepares for its morning orevening ramp up, or during the nighttimes when supply from must-takeresources exceeds demand. Theseeffects have given rise to a newphenomenon referred to as a ‘duckcurve’ because the shape of net loadresembles the body of a duck in whichthe system is at the risk of over generationat the lowest point of duck belly (seechart. Additionally, when there are a lotof renewables on line, and thus a limitednumber of flexible power plants, thesystem has a limited frequencyresponse – a characteristic that iscrucial for the power system to recoverfrom faults (for example, sudden failureof a massive power plant).

In order to operate reliably under theseconditions, new practices in planningand operation need to be introduced.Traditional reliability metrics need to besupplemented by variable generationintegration studies. An example of suchnew metrics is the Insufficient RampingResource Expectation (IRRE); this cancomplement generation adequacymetrics (Loss of Load Expectationand other relevant indices) in orderto assess whether planned capacityallows the system to respond to short-term changes in the net load. Suchoperational changes pave the way

for the power systems of the future inwhich:

resources react quickly and meetexpected operating levels for adefined period of time,

ramp directions can change rapidly,

energy can be stored or its usemodified,

power plants are able to start andstop multiple times per day,

operators can make accurateforecasts of their operating capability.

Flexibility services can be providedby various resources such as: fastramping thermal generations (forexample, open cycle gas turbines),storage, interconnection, demandresponse, and even curtailmentof renewables. The adequacy oftransmission and distribution networks,in terms of having no bottlenecks andsufficient capacity, is also crucial toenable flexibility.

Market and regulation

Such changes in traditional operationand planning of the power system arenecessary but not sufficient to enablea flexible power system. There is also aneed for institutional changes, in termsof market (and product) design and

California Independent System Operator (CAISO) duck curveSource: ‘What the duck curve tells us about managing a green grid’, California ISO Fast Facts.



regulation. These changes are requiredto ensure not only that the currentresource owners have the incentive tooffer their flexibility services, but alsothat the market incentivizes new entrantsto meet the flexibility requirements ofthe system. Many of the conventionalgeneration sources have not beendesigned to take on the ramping andcycling duties that are now required,and such operational modes increasethe of cost of their wear and tear andfuel consumption. Therefore, a market(or regulation) that values flexibility isrequired to enable these generators tocover the costs of operating in a flexiblemode when needed.

In the author’s view, there is a range ofmeasures that can be taken in orderto align electricity markets with therequirements of a flexible power system.These include: modification of thecurrent markets to make efficient use ofavailable flexible resources, designingexplicit incentives for flexibility services,and defining tradable flexibility products.

‘IN MANY POWER SYSTEMS WITH

FLEXIBILITY ISSUES, ELECTRICITY

MARKETS ARE NOT POSITIONED TO

MAKE EFFICIENT USE OF AVAILABLE

RESOURCES.’

In many power systems with flexibilityissues, electricity markets are notpositioned to make efficient use ofavailable resources. For example, insome US regions where sub-hourlymarkets do not exist, normal short-term variation of net load is met byfrequency regulation services – utilizingthe fastest, but most expensive, flexibleresources. However, in other regions(typically those with an ISO) suchvariations are managed at a muchlower cost through procurement insub-hourly markets. The presence ofsub-hourly markets provides a uniqueopportunity for the system operator toexploit the inherent flexibility of a large

fleet of generation sources that canalter their aggregate output rapidly.

The definition of standardized tradableflexibility products is also an importantpart of the game. For instance, inorder to help the system and usedispatchable flexibility, the CaliforniaISO is working on a proposal toincorporate a new product, calledflexible ramping products (FRP). FRPsoffer a five-minute ramping capabilitythat will be dispatched to meet five-minute to five-minute variations in netsystem demand. In this context, the netsystem demand is defined as the loadplus export minus the schedules ofall resources that are not five minute-dispatchable, including renewables,imports, and self-scheduled resources.

Another example in line with suchmodifications in electricity marketsis the UK review of balancingmarkets and changes to cash-outarrangements, which aims to makethem reflect marginal rather thanaverage costs. Sharpening cash-outprices provides incentives for themarket players to offer, invest in, orsecure flexible resources to balancetheir position at the time of systemstress. On top of that, the market forflexibility services can be regionalrather than national. This is specificallyrelevant to the EU region, in whichshort-term balancing markets canbecome integrated where there issufficient interconnection capacityavailable (and will provide incentivesfor investment in interconnections).This promotes efficient procurementof flexibility services because, withthe increase in the market size, moreoptions become available to nationalsystem operators.

In addition to efficiency, anotherobjective of a flexibility market isto ensure the sufficiency of flexibleresources. This requires explicitincentives for adding flexible resources,since conventional capacity markets are

not a substitute for the flexibility market.This is simply because the standarddefinition of resource adequacy doesnot necessarily imply flexibility: asystem can be adequate in terms ofresources available to meet seasonalpattern of load profile (capacity exceedsdemand with a margin at all time) butlack the operational flexibility to dealwith within-day fluctuations of net load.Furthermore, when capacity marketsdo not explicitly value flexibility, theoutcome of procurement favours low-cost resources that lack the capabilityto respond to short-term variation of netload (the UK capacity market is a goodexample for this). The bottom line is thatflexibility is a component of scarcity;thus a single-product capacity marketsuffers from the very same ‘missing’scarcity distortion that it was intendedto address in the first place. Therefore,the current capacity markets in the EUregion should be designed so that bothcapacity and operational capability ofresources (and maybe also carbonemission reduction) are valued.

‘…DEMAND RESPONSE IS A GOOD

EXAMPLE OF HOW REGULATION CAN

AFFECT THE FLEXIBILITY OF A POWER

SYSTEM.’

On the regulatory side, demandresponse is a good example of howregulation can affect the flexibilityof a power system. For instance,the rules regarding the nature ofdemand response (equal treatmentof it with other resources and types ofmarkets in which demand responsecan participate) are important forutilization of this resource. In someUS regions, demand response isallowed to participate in the ancillaryservice markets; this is an attractivearrangement for aggregators ofdemand response services. Sucharrangements are also happening inthe UK, which is trying to integrate thedemand side in its balancing service;



for example, under the FrequencyControl Demand Management (FCDM)scheme, frequency response isprovided through automatic interruptionof contracted consumers when thesystem frequency transgresses the lowfrequency relay setting on site. The UKNational Grid is also trying to utilizeslower-responding demand responsefor load following services under theShort Term Operating Reserve (STOR).

Conclusions

The widespread deployment of

intermittent renewable generation in

several countries has led to an increase

in the relative and absolute magnitude

of power generation with random

output. This deployment constitutes

a major paradigm shift in grid

management, both at an operational

level and also in terms of market design

and regulation. In addition to traditionalresource adequacy metrics we neednew methodologies, in order to quantifyand assess the technically availableoperational flexibility of the powersystem to be utilized for planning andoperation. Also, we need to ensure thatelectricity markets and regulations notonly provide incentives for investmentin flexible resources, but also makeefficient use of available resources.

Market integration with energy-only markets and renewables: lessons ofexperience from AustraliaRabindra Nepal and Tooraj Jamasb

The establishment of wholesaleelectricity markets has been oneof the hallmarks of the market-oriented electricity sector reformsand restructuring process that startedin the early 1990s. The standardmodel of liberalization included theestablishment of wholesale and retailcompetition, vertical separation of thedistinct generation, network, and retailactivities, and incentive regulation ofelectricity networks. A fuller physicaland financial integration of theseparate national or regional electricitymarkets was then required to enable adeepening of competition. For examplein the EU, the process of creating acommon and integrated wholesalemarket for electricity that started inthe second half of the 1990s, remainsa work in progress, while Australiahas, since 1998, focused on creatingan efficient and integrated NationalElectricity Market (NEM).

‘THE ESTABLISHMENT OF WHOLESALE

ELECTRICITY MARKETS HAS BEEN ONE

OF THE HALLMARKS OF THE MARKET-

ORIENTED ELECTRICITY SECTOR

REFORMS…’

This article draws upon the NEM’s

experience with market integration,renewable energy, and networkregulation, and merges the findingsfrom recent studies: ‘Testing forMarket Integration in the AustralianNational Electricity Market’, a paperforthcoming in the Energy Journalby Rabindra Nepal and John Foster;‘Network Regulation and RegulatoryInstitutional Reform: Revisiting the Caseof Australia’, an article in Energy Policyfrom October 2014 by Rabindra Nepaland Tooraj Jamasb; and ‘ElectricityNetworks Privatization in Australia: AnOverview of the Debate’, an article inEconomic Analysis and Policy fromDecember 2015 by Rabindra Nepaland John Foster.

We discuss the context within whichrenewable energy development andmarket integration can be implementedas complementary policies in energy-only markets, citing Australia as aspecific case, but drawing lessonsfor the EU. The NEM provides aninteresting case study, since it is oneof the most transparent energy-onlywholesale electricity markets and it islocated in an island economy, whileit is also poised for greater uptake ofrenewable energy. The lessons from theNEM can be important for other regions

such as the EU, which is predominantlyan energy-only market moving towardsgreater integration of renewables andelectricity markets.

The next section of the article providesan overview of the structure andorganization of the NEM; followingthis, the facts and underlyingreasoning behind the idea that thedevelopment of renewable energyand electricity market integration canbe complementary policy objectivesin Australia, are presented. We thenconclude by highlighting the broaderimplications for other energy-onlymarkets, especially in the EU.

The National Electricity Market (NEM)

The NEM was established in 1998 inresponse to the overall restructuringof the Australian electricity sector. TheNEM is a gross pool arrangement forwholesale electricity trade in Australiaand operates a deregulated marketin the physically interconnected, butseparate, regions of New South Wales(NSW), Victoria (Vic), Queensland(QLD), the Australian Capital Territory(ACT), South Australia (SA), andTasmania (TAS). Tasmania joined theNEM in 2005. Exchanges between



electricity generators and consumerstake place in a spot market through acentrally coordinated dispatch wherethe output bids from all generatorsare aggregated and instantaneouslyscheduled to meet demand. A dispatch(or spot) price is determined every fiveminutes and six dispatch prices areaveraged every half hour to determinethe spot price for each trading intervalin each of the regions of the NEM.

The dispatch price is the energy-only price and does not contain anycomponents for capacity such as‘capacity payments’. The NationalElectricity Rules (the rules) set amaximum spot price (market price cap)of 13,100 AUD/MWh and a minimumspot price (market floor price) of–1000 AUD/MWh for the financial year2013/14. The market price cap preventswholesale spot prices from rising toosharply during extreme peak loads orat times of reduced baseload capacity.The negative market floor price allowsgenerators to pay to stay online whenthe cost of staying online is lower thanthe cost of shutting down and re-startingtheir plants. Hence, the minimum spotprice guarantees dispatch by bidding atnegative prices when a generator is toocostly to turn off.

‘THE GENERATION MIX IN THE NEM IS

DOMINATED BY LOW-COST BASELOAD

GENERATION.’

The generation mix in the NEM isdominated by low-cost baseloadgeneration. Queensland, which hasinstalled capacity that amply exceedsthe region’s peak electricity demand,is a net exporter in the NEM. Victoriaalso benefits significantly from low-costbaseload capacity, making it also anet exporter of electricity. New SouthWales is a net importer of electricityand has limited peaking capacity attimes of high demand. NSW mostlyrelies on local baseload generation.South Australia relies heavily on

electricity imports, although newinvestment in wind power has reducedits import dependency since 2005/06.For example, the registered capacityof wind and solar in SA in 2013 was 30per cent of total capacity, while 2,987MW of wind capacity will be added(equivalent to 61 per cent of the totalgeneration mix) by 2023. Tasmaniais also a net importer of electricity,although it became a net exporterin 2011/12 for the first time since itsinterconnection with the NEM, dueto greater water availability and theinstallation of new gas-fired generation.

The Australian Energy Regulator (AER)determines the maximum revenue(revenue caps) that the transmissionand distribution companies can recoverfrom the network users in Queensland,Australian Capital Territory, andTasmania. Likewise, the AER sets themaximum network tariffs that distributorscan charge consumers (price caps)in New South Wales, Victoria, andSouth Australia. The transmission anddistribution networks remain state-owned in Tasmania, New South Wales,Queensland, and part of the AustralianCapital Territory. Victoria and SouthAustralia have fully privatized electricitygeneration. Victoria corporatized andprivatized its electricity networks between1995 and 1999 and both the transmissionand five distribution networks are nowin private ownership. South Australiahas privately owned transmissionnetworks, while the distribution networkis leased to private interests.

There are six operationalinterconnectors among the fiveelectrically connected states in theNEM. There are two interconnectorsoperating between NSW and QLD, andtwo between SA and VIC, while VIC–TAS and NSW–VIC are each connectedby a single interconnector. There isno direct interconnection betweenQLD–SA and NSW–SA. The existinginterconnectors largely follow the stateboundaries, covering a distance of

more than 5000 kilometres, runningfrom Port Douglas in Queenslandto Port Lincoln in South Australia.Hence, NEM is one of the longestinterconnected power systems inthe world. Geographical constraintsas an island economy have ledto the infeasibility of cross-borderinterconnections, to date, in the NEM.

Market integration outcomes

Despite more than 14 years havingpassed since the establishment ofthe NEM, wholesale electricity pricedifferences persist across the Australianstates. Average daily prices are lowestin VIC and QLD, while SA has thehighest average price, followed by TASand NSW. However, the average dailywholesale price differences betweenSA and VIC are the lowest among thephysically interconnected states. Thepersistent differences in wholesale pricescan be attributed to the presence ofnetwork constraints across the regionalinterconnectors; this impedes thewholesale market integration processin the NEM. The Australian ProductivityCommission has expressed concernsabout under-investment in transmissionnetworks and in regional interconnectors.In response, the AER has, in the past,recognized the significance of congestioncosts in the NEM and has allowed moreinvestments in the transmission network.

Network constraints occur due to physicallimits to the network’s transfer capability.Network congestion can segment themarket and increase the wholesaleelectricity price by displacing low-priced generation with more expensivegeneration. Congestion can also leadto market power in the segments ofthe market. Finally, congestion alsopromotes inefficient electricity tradeflows between the regions, as electricitycannot be stored and ‘demandand supply’ have to be balancedin real time. The existing networkconstraints, due to underinvestment in



interconnectors, act as a barrier to thewider, and much expected, developmentof renewable resources, such as wind.

‘THE LACK OF ADEQUATE

INTERCONNECTION CAN PREVENT THE

DEVELOPMENT OF NEW WIND POWER

IN THE RESOURCE-RICH REGIONS’

Australia’s wind resources are mostlyconcentrated in the regions with thelowest electricity demand, theseinclude South Australia, Tasmania, andVictoria. Queensland and New SouthWales exhibit high demand for electricity,but have a low concentration of windresources. For example, South Australiacan reach a wind penetration (percentageof average generation) of almost 70 percent, followed by Tasmania at 50 percent, while Queensland and New SouthWales have wind penetrations of around1 per cent and 11 per cent, respectively.The lack of adequate interconnection canprevent the development of new windpower in the resource-rich regions, andlead to curtailment of existing capacity.The costs in terms of lost revenue tothe wind power plants as a result ofcurtailment are large, and hamperthe country in its attempts to meet itsnational renewable energy targets.

Furthermore, wind projects are graduallymoving to less congested parts of thenetworks across Queensland and NewSouth Wales, but these areas enjoylower quality wind resources, andwill eventually face higher costs. Theexpansion of interconnectors, togetherwith increased export capacity fromlow-demand regions (with high windpotential) to high-demand regions, isimportant in order to reap the security ofsupply and sustainability benefits fromwind generation in the NEM. Improvingmarket integration by increasing thecross-border power flow will also leadto benefits through efficiency gains,both allocative and productive, as wellas through dynamic efficiency gains,because a well interconnected market will

facilitate the optimization of investmentsin both generation and transmission overtime across the NEM market.

Conclusions and policy implications

The most important factor for marketintegration is to connect regionalelectricity systems, physically andadequately, through a transmission gridand interconnectors. The transmissioncapacity and prices will then determinethe volume of trade between thedifferent regions.

The EU is striving to create anintegrated electricity market in Europe,while also aiming to significantlyincrease the share of its energy fromrenewables. However, as the EU hasidentified, the European transmissionand distribution networks need to beadapted and extended to facilitatepower flows from generation source toend users across borders. Achievingthis objective will require substantialinvestments. The island economieswith isolated electricity markets – inNorthern Ireland and the Republic ofIreland – are also aiming to increaseintegration with the continentalEuropean electricity markets under theEU ‘Target Electricity Model 2014’. Inaddition, the Republic of Ireland hasa target to generate 40 per cent of itselectricity from renewables by 2020.

The NEM experience with marketintegration suggests that harmonizationof regulatory and institutionalframeworks and electricity marketregulations can be coupled with privateownership of assets, to improve marketintegration across energy-only marketspoised for a large intake of renewablesin the wholesale trade. The case ofNEM suggests that a solid regulatoryframework could be a pre-requisite forthe necessary infrastructure investmentsto take place in time for 2020. Forexample, the AER recognized thesignificance of congestion costs acrossthe regional interconnectors and allowed

higher transmission investments inregulatory decisions in 2009.

The regulatory test for transmissionexpansion and network planningin the NEM is based on identifyinginvestment options that maximizethe net economic benefits. However,the EU currently has 28 differentnational regulatory frameworks. Afragmented regulatory system basedon uncoordinated national policies canbecome an obstacle to the formation ofan internal electricity market. Achievingan integrated European market ischallenging, given the lack of adequateinterconnections, and inconsistencyin market design and rules among themember countries, while aiming toincrease the share of renewable energy.

‘THE EU MEMBER STATES NEED TO

IMPROVE THE HARMONIZATION OF

THEIR REGULATIONS AND INCREASE

COOPERATION…’

The Australian experience revealsthat the large-scale development ofrenewable energy and the integrationof regional electricity markets can becomplementary; these policies do notnecessarily conflict with each otherunder adequate transmission capabilityin energy-only markets. The EU canimprove the number and capacity ofinterconnections in order to achievea more resilient energy-only marketand to implement the Energy Union(currently uncertain due to inadequateinvestments in electricity networks).The regulatory framework for wholesalemarkets and networks will be importantfor facilitating trade across theinterconnectors; it will thereby improvemarket integration among energy-onlymarkets with a high share of renewableenergy. The EU member states needto improve the harmonization of theirregulations and increase cooperation,in order to attract economicallybeneficial investments and achieve theEnergy Union.



Reforming European electricity markets for renewablesRichard Green and Goran Strbac

If the European Union is to meet its2030 target and get 27 per cent of itsenergy from renewable sources, it islikely that roughly half of its electricitywill have to be from renewables. In2013, renewable sources provided27.7 per cent of electricity generationin the EU-28, including 12.3 per centfrom hydro, 7.2 per cent from wind, and2.6 per cent from solar power. A largeincrease in solar and wind capacityseems inevitable if the 2030 target isto be met, but this will pose significantchallenges for the electricity system.More transmission capacity is planned,but the existing infrastructure needsto be used better, and that requires anappropriate market design. We arguehere that more cross-border tradingwill be needed – of energy in the longterm, and of balancing services (suchas reserve) in the short term – andsuggest changes that would facilitatethis. The cost of expanding Europe’srenewable generation could besignificantly reduced if coordinationacross the EU replaced the presentpatchwork of national approaches.

European integration and transmissioncapacity

Europe’s underused transmissioncapacity imposes significantopportunity costs on its people.Transmission currently supportsenergy arbitrage in day-aheadscheduling, buying in currently cheapmarkets and selling in more expensiveones, but it could also supportcapacity sharing and more efficientbalancing in real time.

The benefits of moving from the currentmember state-centric market designto one which is pan Europe-widehave been quantified by modellingat Imperial College, which providedthe basis of two 2013 reports to theEuropean Commission: one detailingthe benefits of an integrated Europeanenergy market, and the other givingan impact assessment on a Europeanelectricity balancing market. Thatanalysis demonstrated that the benefitsof fully integrating EU energy andcapacity markets would be EUR12–40billion/year and EUR7–10 billion/yearrespectively by 2030, while integrationof the EU balancing market wouldsave an additional EUR3–5 billion/year. These savings go far beyond theEUR2.5–4 billion/year that the EU hassaved through its existing measures tointegrate its electricity markets throughday-ahead energy arbitrage. Thepotential gains would be even greaterif transmission capacity was used tosupport a more rational deployment ofrenewable generation.

Comparative advantage – significance oftransmission

The policies of most EU governmentshave had the effect of spreading windand solar power across the continent,making support available to a widerange of renewable generators. TheUK government set feed-in tariffs atlevels that encouraged the installationof 8.3 GW of solar PV capacity (toOctober 2015), even in the cloudy,northerly British Isles, while the Germangovernment has supported 34.7 GWof wind capacity, even with a 19 percent average load factor that is onlytwo-thirds of the average for the UKand Ireland. In other words, nationalpolicies have ignored the principle ofcomparative advantage.

In a paper forthcoming in The EnergyJournal, working with Iain Staffell andDanny Pudjianto, we have calculatedthe benefits and costs of a morerational deployment of renewablecapacity, one that would site a higherproportion of generation in areaswhere it will have high load factors. Wefind that by moving wind generationtowards the windier areas of north-westEurope, it would be possible to get thesame level of output from 15 per centless capacity. Concentrating solar PVgeneration towards the south wouldsave 8 per cent of installed capacity.Overall, this would save EUR19 billiona year in interest and depreciationcosts. However, power flows across thecontinent would increase significantly,since a concentrated portfolio ofrenewable generators would producestronger peaks and troughs in outputthan a dispersed one. Our modellingsuggests that an additional EUR4billion a year would be needed tofinance additional transmission linesand peaking generation, for times whenthe local renewable output was low andeven the new lines were congested.The net gain is therefore considerable,but there are significant challenges toovercome before such a coordinateddeployment could be possible.

European energy market design

Many of these challenges are inthe areas covered by the EuropeanCommission’s July 2015 consultationon a new energy market design.Electricity markets in Europe largelyfollow a common pattern; much tradingis bilateral and concluded some timein advance, so that both generatorsand retailers can fix a price for most oftheir sales and purchases. There is avoluntary day-ahead auction run by anindependent power exchange, which

‘…THE EXISTING INFRASTRUCTURE

NEEDS TO BE USED BETTER, AND THAT

REQUIRES AN APPROPRIATE MARKET

DESIGN.’



may cover one country or a small groupof neighbouring countries. Adjustmentsclose to the time of delivery are madeby the local transmission systemoperator, trading bilaterally withgenerators and a few consumers (ortheir aggregators) able to offer demandreductions at short notice.

‘ENCOURAGED BY THE EUROPEAN

COMMISSION, MARKET COUPLING HAS

GRADUALLY SPREAD ACROSS NORTH-

WEST EUROPE…’

One spot market, Nord Pool, hasbeen multinational since 1996, anduses a market splitting algorithm toset different prices on either side of azonal or national border if transmissioncapacity is less than that required forthe flows that would equalize prices.France, Belgium, and the Netherlandsstarted market coupling in 2006, takingaccount of exports and imports intheir national auctions, and settingtheir level to equalize prices acrossborders if possible, and exhaust theavailable transmission capacity ifnot. Encouraged by the EuropeanCommission, market coupling hasgradually spread across north-westEurope, so that 19 countries hadcoupled their day-ahead marketsby February 2015. If there is enoughtransmission capacity, an area fromPortugal to Great Britain, to Norway,Finland, and the Baltic States, and fromGermany via Austria and Slovenia toItaly could have a single day-aheadprice.

Although day-ahead markets may bethe most visible sign of a liberalizedelectricity industry, most electricitytrading actually takes place on alonger timescale, while secure systemoperation depends on very short-termdecisions. Long-term contracts allowgenerators and retailers to fix a pricefor most of their expected outputand purchases, giving them financialsecurity, even though the day-ahead

markets allow them to adjust theirpositions to reflect actual patterns ofdemand and plant availability. Thisability to adjust positions is increasinglyimportant as the share of intermittentrenewables rises, since their outputscannot be accurately predicted on thetimescale of long-term trading.

If Europe is to concentrate its renewablegenerators in areas of comparativeadvantage, however, long-term salescontracts will certainly be required: theconsumers supporting (still relativelyexpensive) generators abroad will wantto know that they are getting some powerin return for their support. This requiresa mechanism for delivery, and long-termaccess to capacity on cross-borderinterconnectors has been made availablethrough auctions for many years. Thereare two problems relating to the use ofcontracts for physical capacity.

1 In the past, there were fears that acontract holder might withholdcapacity from the market in order towiden price differentials between thetwo ends of the interconnector. Thisproblem can be resolved with a ‘use itor lose it’ rule that returns any unusedcapacity to the market if the contractholder has not scheduled the flows towhich they are entitled.

2 The second problem, morefundamental, is that physical contractsdo not reflect the physics of the grid.Power flows can be netted off eachother. If 2 GW is sold from France tothe UK, this would appear to use thefull capacity of the cross-Channel link,but if a second deal sends 1 GW inthe opposite direction, a third tradebecomes possible, giving a grosscontracted flow from France to the UKof 3 GW. It is highly unlikely that wecould find a watertight way to issuephysical contracts for 3 GW of powerto flow over a 2 GW interconnector,because the third GW depends on acounter flow that might not happen inpractice.

Financial transmission rights (FTRs)do not have the same restrictions asphysical contracts. They offer priceinsurance for a cross-border tradeby paying the buyer the differencein prices between two points onthe system. As a purely financialcontract, there is in principle no limitto the number of FTRs that could bebought and sold. In practice, thereis a natural hedge for each FTR, andthat is the revenue received by theowners of an interconnector whenpower flows across it from a lower- to ahigher-price area. The capacity of thetransmission system determines thenet volume of FTRs that can receivethis natural hedge – at most 2 GW ineither direction in the case of the linementioned above. However, a netvolume of 2 GW is consistent with2 GW of contracts in one direction, or3 GW in one direction and 1 GW in theother, and so on. If the holders of 3GW of FTRs from France to Englandare receiving a payment becausethe power price is higher in England,then the holders of the FTRs fromEngland to France are making thesame payment (per MW). The net resultis that the transmission companiesmake the payment for 2 GW of FTRs,just matching the capacity thatthey would normally have available.Absent bankruptcy, the holders of theEngland–France FTRs are guaranteedto provide a financial counter flowin a way that cannot be ensured forphysical contracts.

Creating long-term FTRs might also bea way of financing the extra investmentin transmission assets that will beneeded to accommodate the morevolatile power flows as renewablecapacity rises. Companies expectingto trade power over long distanceswill pay the companies developingtransmission assets for the FTRs thatthe traders need to lock in prices; thecombination of the FTR payments andthe price differences that they hedge



gives the transmission owners a morestable revenue stream, independent ofday-to-day fluctuations in fuel prices orrenewable outputs.

FTRs would help create a single electricitymarket that operates on long timescalesbut the short-term markets that are crucialfor the secure delivery of power remainlargely national. While system operatorsrely solely on plants and consumerswithin their own borders to providebalancing services, they are likely to missalternatives that offer better value. In theearly days of the New Electricity TradingArrangements in England and Wales in2001, the system operator had to sethigh imbalance prices because of thecost of buying power from a relativelysmall number of flexible generators insidethe market. As soon as the operatorrealized that flows from France andScotland could be adjusted by similaramounts within the required timescales,the presence of a much cheaper optionhelped prices to fall dramatically.

Balancing mechanisms that allowcross-border participation requirethe effective exchange of informationon short timescales. They also needtransmission capacity to be available,and we do not currently have amechanism that allows for this. If cross-border capacity is entirely allocatedat the day-ahead stage, there will benone available for the country importingenergy to also import balancingservices from its neighbour.

Case for Transmission System Operators

North American power markets jointlyoptimize energy and reserve decisions,for they have learned that separateoptimization increases costs forconsumers. Since the markets are runby Independent (transmission) SystemOperators, they take the constraints onthe transmission system into accountwhen they do this.

Transmission capacity in Europe iscurrently allocated by power exchangestrading energy, while separateinstitutions deal with reserve andbalancing. That implies that all thetransmission capacity is allocated toenergy arbitrage, which is unlikely tobe the optimal outcome. The EuropeanCommission sought views on whethercross-border Transmission SystemOperators should be encouraged:promoting more efficient balancingis one area in which they wouldhave an advantage over the currentarrangements. In principle, and alreadysometimes in practice, the operatorin one country can acquire balancingservices from its neighbour, but theextra stages of decision making andcommunication involved are likely toreduce the benefits of doing so.

It is a sign of progress that theEuropean Commission thinks theestablishment of cross-borderTransmission System Operators isworth consulting on: in the past,

national electricity markets werefiercely separate. National renewableenergy policies are still uncoordinated,and concentrating renewables inmore favourable locations wouldbring important savings. Even withthe current capacity mix, the gainsfrom greater trade have becomeapparent, and will increase as theshare of wind and solar generationcontinues to rise. The existing linkagesbetween day-ahead markets needto be supplemented with greatercoordination on longer and shortertimescales. Cross-border trading oflong-term contracts for energy wouldbe facilitated by the use of FinancialTransmission Rights that hedge pricedifferences between areas, whilerespecting the laws of physics in a waythat past commercial arrangements forlong-distance power flows have not.Short-term balancing costs could bereduced if system operators sometimesbought balancing services and reservefrom their neighbours. To integratebalancing between countries, however,we may have to integrate energy andreserve markets inside each country.American experience shows thatthis can be done, but the politicalchallenges in creating a good marketdesign should not be underestimated.

Supporting renewable generation in the UKDavid M. Newbery

Introduction

The European Commission’s proposed

Energy Union Package (introduced in

the document subtitled ‘A Framework

Strategy for a Resilient Energy Union

with a Forward-Looking ClimateChange Policy’ dated 25 February2015) proposes integrating electricityfrom renewable energy sources(RES-E) into the market. It aims tomove away from the unresponsive

standard feed-in tariffs (FiTs), replacingthem by Premium FiTs (PFiTs), whichpay a premium on the market price butrequire generators to take responsibilityfor selling and balancing their power.This article examines the logic,

‘…CONCENTRATING RENEWABLES IN

MORE FAVOURABLE LOCATIONS WOULD

BRING IMPORTANT SAVINGS.’



drawbacks, and possible resolutions ofthis proposal for future UK renewablessupport.

History of UK renewables support

Britain has tried almost every formof renewables support since the firstNon-Fossil Fuel Obligation auctionsin the 1990s and so is a usefultestbed for the efficacy of differingforms of renewables support. Asthe name suggests, the first formof support was a series of auctionsfor the prices to be paid for variouscategories of non-fossil fuel (whichwas extended to other non-fossil fuelsfrom nuclear power under pressurefrom DG COMP). The early auctionsdemonstrated their competitive effectby driving down costs (or the auctionprices), but either the winner’s curse,or the lack of any penalty for multiplebids between which the winner couldchoose in light of subsequent planningapplication success, led to a fall in therate at which auction winners deliveredcommissioned plant. Inflexibilities inproject definition (size, exact location)also meant that small changesrendered the initial contract invalid.

‘BRITAIN HAS TRIED ALMOST EVERY

FORM OF RENEWABLES SUPPORT

SINCE THE FIRST NON-FOSSIL FUEL

OBLIGATION AUCTIONS IN THE 1990S…’

These objections could surely havebeen overcome, but a change inGovernment in 1997 from Conservativeto Labour resulted in a reconsideration,and with it the Utilities Act 2000 thatplaced an obligation on electricitysupply companies to source aspecified share of their sales fromrenewable generation, or pay a pre-setpenalty for any shortfall. Companiesproducing RES-E were awardeda specified number of RenewableObligation Certificates (ROCs) foreach MWh they produced; however,

they were responsible for sellingthe electricity and dealing with anyimbalances, in contrast to the lessrisky FiT schemes which required theSystem Operator to handle all off-take,and under which the developer waspaid a set price for metered output. Thevalue of ROCs is determined by supplyand demand and the penalty price fornot delivering sufficient ROCs to meetthe targets.

One obvious advantage that quantity-based instruments like ROCs enjoyis that the volume of RES-E to besupported can be limited, and hencethe impact on the budget is alsopredictable and controllable. In thecase of ROCs, the volume required tobe bought by the utilities is set eachyear, usually with some ‘headroom’that is an estimated uplift on thepredicted capacity during the year.The excess demand from retailers’penalty payments produces additionalrevenue that is returned to increasethe value of the ROCs. By reducingthe headroom, or if developers areunusually successful and exceed thepredicted amount, the market price forROCs will fall and discourage furtherexpansion. In contrast, a fixed FiT withno cap on the number that can beissued risks excessive demand andexcessive budgetary cost. While winddevelopments take years to proceedfrom the initial idea through planningto grid connection agreements andconstruction, and hence can bepredicted well in advance, solar PVpanels can be ordered and installedwithin weeks. Spain and Italy (and theUK) have demonstrated that if the FiTis set too high compared to the rapidlyfalling cost of PV, installation ratescan accelerate and create such fiscalpressures that the support system cancollapse.

One other important advantage thatROCs or PFiTs have is that they cangive better locational signals than theclassic FiT, which normally pays the

same amount regardless of location.In the UK, wind developers have topay locational Transmission NetworkUse of System charges, which varyvery significantly across the country.If electricity were locationally spotpriced, as in many parts of the USA,the temporal and spatial signalswould provide additional efficientlocation signals, with the premiumpayment remaining as a support tothe high capital cost. The limitation tothese locational signals is that if thesame premium is given regardless oflocation, the benefit of generating in ahigh wind area with a higher capacityfactor will be unnecessarily amplified,distorting location decisions to suchlocations and providing excess rent tothe developers there.

‘… THE IDEAL FORM OF SUPPORT

WOULD TARGET THE CAPITAL COST,

NOT THE OPERATING COSTS.’

As the aim of the support is toencourage deployment of high capitalcost plant to drive down future costs,the ideal form of support would targetthe capital cost, not the operatingcosts. Germany has a very simplesolution which largely achieves this bypaying support for a specified numberof MWh/MW capacity, so windier areasreceive only slightly more support thanless windy areas (by receiving it morequickly). This has the advantage ofreducing excessive rent to high windareas, and effectively treating theelectricity produced as of equivalentvalue to any other electricity – correctly,as the electrons are not coloured greyor green depending on source.

The downside to the benefit ofcontrolling the quantity and hence thefiscal cost of support is that developersfind it hard to predict their futurerevenue. With ROCs there is a doubleuncertainty: the future value of ROCsand also the future price of electricity.Further, the price of electricity in the



UK is set by either coal or gas (and theprice of carbon), as fossil plant remainsat the margin for all but a few hours peryear. As a result fossil plant enjoys anatural hedge while low-carbon plantwith low variable costs is fully exposedto the very volatile wholesale price.

The result of the higher risk facingRES-E developers in the UK hasbeen a slower uptake than might beexpected, given that the UK has farbetter wind resources than Germany.In 2004 the UK generated about 8per cent of the amount that Germanygenerated from wind, but by 2011 theproportion had risen to more than 30per cent, reflecting the UK’s higher rateof growth as it started from a very lowbase (figures taken from the Eurostatwebsite).

In response to concerns about the highcost of supporting RES-E, combinedwith its lagging performance, as wellas growing concerns over security ofsupply, the UK Government passedthe Energy Act 2013, which launchedthe Electricity Market Reform (EMR).EMR was intended to meet the EU RESand climate change targets at lowercost while maintaining reliability. Incontrast to the Energy Union Package,it aims to replace PFiTs with Contracts-for-Differences (CfDs) for RES-E,

closer to the classic FiT, as the wayto lower the cost of RES-E support.The Department for Energy & ClimateChange (DECC) initially set the strikeprices for these CfDs on the basis of anassumed hurdle rate, which the Panelof Technical Experts, commenting in a2013 report on the delivery of the EMR,criticized as being too high. Instead,the Panel suggested that auctionswere a better way of eliciting the priceat which developers would be willingto invest, and under pressure fromDG-COMP again, DECC announcedan auction for RES-E CfDs on 1September 2014. The results of thatauction are given in Table 1.

It is clear that the auction pricesare considerably less than theadministratively set strike prices and itis simple to estimate the reduction inthe implied hurdle rate by comparingthe administered strike price with theauction price. Given the auction prices,the administered strike prices, andassumptions about the cost of windturbines, their capacity factors, andoperating costs it is straightforwardto compute the difference in theinternal rates of return between thetwo alternative income streams.The differences are large at 2.2–3.4per cent real (depending on these

parameters and whether any credit isgiven for post-contract operation at theunsubsidized price).

The Energy Union Package

Just after the UK’s first CfD auctionin February 2015, the Energy UnionPackage was launched, stating that:

… renewable production needsto be supported through market-based schemes that addressmarket failures, ensure cost-effectiveness and avoidovercompensation or distortion.Low-cost financing for capitalintensive renewables dependson having a stable investmentframework that reducesregulatory risk.’ (Energy UnionPackage, 2015).

This Commission proposal appears topay more attention to the advantagesof PFiTs mentioned above than to theirimpact on risk and financing costs,and would seem to reverse the logic,painfully learned in the UK, of movingfrom PFiTs to FiTs with their revenueguarantee and hence reduced risk andcost. German, Danish, Spanish, andItalian case studies (reported in twoarticles in the journal Energy Policy:‘Environmental policies and risk finance

CfD auction allocation: round 1

TechnologyAdminprice

Lowestclearing

price 2015/16 2016/17 2017/18 2018/19

Totalcapacity

(MW)Advanced conversiontechnologies

GBP/MWh 140 114.39 119.89 114.39MW 36 26 62

Energy from waste withcombined heat and power

GBP/MWh 80 80 80.00MW 94.75 94.75

Offshore wind GBP/MWh 140 114.39 119.89 114.39MW 714 448 1162

Onshore wind GBP/MWh 95 79.23 79.23 79.99 82.50MW 45 77.5 626.05 748.55

Solar PV GBP/MWh 120 50.00 50.00 79.23MW 32.88 36.67 69.55

Note: the GBP50 bid for solar PV in 2015/16 was withdrawn.

Source: DECC ‘Contracts for Difference (CfD) Allocation Round One Outcome’, 26 February 2015.



in the green sector: Cross-countryevidence’ by Chiara Criscuolo andCarlo Menon, 2015, and ‘Lessons foreffective renewable electricity policyfrom Denmark, Germany and theUnited Kingdom’ by Judith Lipp, 2007,and also in the EPRG Working Paper1603 ‘Energy subsidies at times ofeconomic crisis: a comparative studyof Italy and Spain’ by Arjun Mahalingamand David Reiner) demonstrate thata well-designed FiT can be costeffective (with suitable degressiontracking falling costs), can deliver rapiddeployment, and encourage the costreductions that are the logic behind theEuropean Commission’s RenewableEnergy Directive.

‘THERE ARE GROWING CONCERNS …

THAT THE UNANTICIPATED MASSIVE

INCREASE IN RES-E PRODUCTION IS

LOWERING WHOLESALE ELECTRICITY

PRICES…’

However, there are growing concernsfrom electricity generating companiesthat the unanticipated massive increasein RES-E production is loweringwholesale electricity prices, particularlyin Germany, which has massive windand solar PV penetration (see, forexample, ‘The impact of wind powergeneration on the electricity price inGermany’ by Janina Ketterer in the

journal Energy Economics, 2014). Thereis concern that excessive generation atsome particular time and place wouldnormally lower prices and discouragefurther investment there, but a fixedFiT would remove that feedback. Onthe other hand, a PFiT that just pays a(normally fixed) premium on the localspot price would provide the necessaryfeedback.

The practical question is how tocombine the advantages of a PFiTwith the risk-reducing properties of aFiT or CfD. The logic of the RenewableEnergy Directive is that it solves the‘club good’ problem of financingdeployment to reap the dynamiceconomies of scale (learning-by-doing), which is primarily more aboutthe design, location, and installationof the RES-E plant, and less about itsoperation (which, if it is mature enoughto warrant mass deployment, shouldprimarily depend on the resource:wind or sun). This would suggestrewarding RES-E for availability ratherthan output, or per MW rather thanper MWh, making renewables just likecapacity in a capacity auction (of thekind successfully implemented as partof the EMR), as the aim would be toidentify the ‘missing money’ neededto justify deployment, while providinga long-term contract for availabilitythat addresses the ‘missing (futures)

market’ problem. See the forthcomingarticle by David Newbery ‘MissingMoney and Missing Markets: Reliability,Capacity Auctions and Interconnectors’in Energy Policy.) To reduce risk further,balancing and other ancillary servicescould be procured competitively bythe System Operator, while the RES-Edeveloper could be offered a cost-reflective contract whose cost would befactored into the auction for capacityavailability. Other aggregators or supplycompanies could offer power purchaseagreements (PPAs) for the meteredoutput, based on a prediction of thelocal wholesale price, further reducingtransaction costs and risks.

Conclusion

The intention behind the EnergyUnion Package – of making RES-Eface more efficient price signals – issound but needs to be reconciledwith a sufficiently stable investmentclimate that allocates risk to thosebest placed to bear it, while providingadequate incentives for efficiency. Thisarticle argues that capacity contractswith suitable PPAs and contracts forancillary and balancing services (withall contracts reflecting efficient marketvalue) are the logical solution andshould support the delivery of low-carbon electricity at least cost.

Centralized or decentralized? Remove first the regulatory barriersIgnacio Pérez-Arriaga and Scott Burger

Regulation matters

Electric power systems are currentlyfacing significant changes as a resultof the deployment of information andcommunication technologies (ICTs),advanced power electronics, anddistributed energy resources (such asgas-fired distributed generation, solar

PV, small- and medium-sized wind

farms, electric vehicles, energy storage,

and demand-side management).

These distributed energy resources

(DERs), unlike ‘traditional’ centralized

generating units, are characterized by

their small capacities (several kilowatts

to several megawatts), their diverse

nature (generation, storage, responsive

demand, or any combination thereof),

and their connection to low and

medium voltage electricity distribution

grids. DERs, if properly integrated, may

have the potential to deliver not only

the valuable electricity services that are

traditionally provided by centralized



generating units, but also new servicesthat are enabled by their distributednature.

‘WHILE DERS ARE CERTAINLY GAINING

TRACTION IN POWER SYSTEMS, THE

DEGREE TO WHICH THEY WILL PROVE

INFLUENTIAL REMAINS AS DIFFICULT TO

PREDICT AS EVER.’

Bold claims have been made – by theCEOs of some of the most importantelectric utilities in the world, byregulatory authorities like those fromthe state of New York or the UK, bythe European Commission, by themost important associations of thepower industry, and by energy thinktanks – announcing the future, evenimminent, disruptiveness of DERs inthe provision of electricity services.Our research group at MIT – currentlyworking on the ‘Utility of the FutureStudy’, in collaboration with the Institutefor Research in Technology (IIT) atComillas University in Madrid (detailsavailable on the website of the MITEnergy Initiative) – has surveyed over200 different kinds of recently createdbusiness models that employ DERsin some way. Something is certainlygoing on, but, while DERs are certainlygaining traction in power systems,the degree to which they will proveinfluential remains as difficult to predictas ever.

From a strict economic viewpoint,the answer to the question of whetherthe future provision of electricityservices will be predominatelycentralized or decentralized, and inwhat likely proportion, will depend onthe characteristics of the services tobe provided and the cost andperformance of the diverse means ofsupplying them. Other factors – suchas customer preferences driven bycultural background, environmentalconcerns, fashion, satisfaction withor animosity towards the incumbent

utility, social pressure, and the impactof neutral or biased information – willalso have a significant influence on theoutcome.

This paper argues that regulation ofthe power sector, as the materializationof national or supranational energypolicy, either correctly designed orflawed, is perhaps the most importantfactor presently influencing thelevel of penetration of DERs in theshort to medium term. The presentregulation is woefully inadequate tomeet the incoming challenges; anin-depth review and correspondingmodification is essential to create aneconomically neutral playing field,enabling centralized and decentralizedresources to compete and collaborateefficiently, while recognizing that theyperform under very different conditions,in terms of size, technology, orlocation in the network, among othercharacteristics. This is an urgent task,since so much is at stake. The irruptionof DERs has added pressure to theneed for regulatory changes to beintroduced (changes that should havebeen made some time ago).

This paper makes the case that goingback to applying the fundamentals ofmicroeconomics to power systems,is the best way of adding value to thecurrent debate about decentralizationand evolution in the provision ofelectricity services. This requires:

1 Reconsidering the definition of‘essential electricity services’;

2 Examining how to compute theprices and regulated charges thatshould apply to every agent in thesystem so that economic efficiency(optimization of social welfare) ismaximized; and

3 Understanding what, if any, is thevalue that aggregation of DERsand any associated businessmodels may bring to the entire powersystem.

Essential electricity services and theirprices and charges

Regardless of the regulatoryframework, only a small number ofmutually exclusive and collectivelyexhaustive electricity services areneeded for the correct (efficient,reliable, and environmentally sound)functioning of a power system: energy(kWh), firm capacity (kW), operatingreserves at several time responselevels, network capacity (kW), voltagesupport, and management ofconstraints and losses. We termthese essential services ‘primaryservices’.

Some of these essential services, likeoperating reserves, may exhibit slightlydifferent definitions under differentconditions; however, the basicconcept does not change. Perhaps inthe future, the physical evolution of thepower sector and its organization willrequire the addition of novel primaryservices to our list; however, wepostulate that the set describedherein is exhaustive given today’sparadigm, and that any potential futureadditions will adhere to the samephilosophy that we describe.

A comprehensive system of economicsignals is needed to allow theanticipated large diversity of agentsto compete and to collaborate in theefficient provision of these services,while maintaining the reliability of thepower system. These economic signalsare prices (for those services providedin competition via markets) andregulated charges (for the regulatedmonopolistic activities); furthermore,these signals act as the power sector’snervous system, efficiently andeffectively communicating, to all theproviders and consumers of electricityservices, the prices and charges thatcorrespond to the particular servicesprovided in their specific situation intime and space.



Prices

Strict application of microeconomicsto the joint mathematical formulationof power system operation andplanning results in prices that are thedual variables – also called shadowprices – of the constraints in theproblem, with the commodities of thecorresponding electricity services beingthe magnitudes on the right hand sideof these constraints. In simpler words,the price is equal to the additionalsystem operation cost of tightening theconstraint by one unit. For instance,the constraint requiring the balanceof demand and generation, at eachnetwork node and at any time, results ina specific energy price at each momentin time at each node: the energy nodalprice or ‘locational marginal price’(LMP). Other constraints emergedue to the system operator’s desireto guarantee that certain essentialservices (such as primary, secondary,and tertiary reserves, or firm capacitytargets, perhaps with some minimumflexibility requirements) are providedin prescribed quantities. There arealso prices for relieving voltage- andcongestion-related networkconstraints.

‘… THE PRICE IS EQUAL TO THE

ADDITIONAL SYSTEM OPERATION COST

OF TIGHTENING THE CONSTRAINT BY

ONE UNIT.’

Regulators may want to imposeadditional constraints – such asemissions or renewables targets, fuelquotas, or banned technologies – allof which result in additional servicesand their associated prices andcharges. Secondary services can bederived from the essential ones, suchas financial products based on theprices of energy or networkcongestion.

LMPs are just energy prices, but theyare a reflection of the diverse costs

of production and demand responseeverywhere, as well as of the impactsof network losses and technicalconstraints. Ideally nodal pricesshould reach every customer, andeven every appliance, but practicalreasons may advise differently.Deciding the optimal granularity – inother words, how far to go withspatial and time differentiation – of allrelevant prices and charges is a majordesign issue for the power sector ofthe future. This is strongly related to thevalue of aggregation that is discussedlater.

ChargesNetwork charges

Regulated network charges arenecessary to recover and to allocateproperly the total network costs.Given the physical and economiccharacteristics of actual networks,this is not possible with just thedifferences between LMPs.Distribution utilities have to applynetwork use of system (DNUoS)charges to recover the total costsof investment, operation, andmaintenance. DNUoS charges canalso signal to network users how theirutilization patterns impact networkcosts.

Regulators are faced with the doublechallenge of: a) recovering networkcosts and b) sending efficienteconomic signals to the networkusers. These signals must enable alevel playing field between electricityservice business models connectedeverywhere in the network to exist.Furthermore, the implementationof network charges must notunnecessarily distort energy prices.

The customary design of DNUoScharges, meant for pure consumingagents in power systems where DERsare considered a minor exception,does not hold anymore. It should beurgently fixed, before more substantial

distortions occur. The new designshould conform to these criteria:

Ignore what happens beyond themeter; network charges should onlybe based on the network locationand the individual profiles of netpower injection or withdrawal.

Assume a well-establishedprocedure exists that designs anefficient or ‘well-adapted’ network foran estimated future pattern ofinjections and withdrawals; thisprocedure should allow theidentification of the underlying costdrivers, such as Connection (thenetwork that is needed to provide a‘basic or minimum’ level of electricityservice to all network users),Capacity (the additional network thatis needed to meet the expected mostdemanding conditions of injectionsand/or withdrawals), and Reliability(the extra network necessary tosatisfy some required quality ofservice standard for all operatingconditions).

The individual network charges willbe determined following a commonmethod that determines thecontribution of each individualutilization profile, at any givenlocation, to each cost driver. Thecosts of any network capacity that isnot directly required by the costdrivers should be socialized.Socialization of network costs is acommon practice today, but a soundjustification is needed.

The network charges will be theamounts of money (per month, forexample) associated with each costdriver and this is how they should bepresented in the electricity bills.Traditional tariffs (USD/kWh, USD/kW,USD/customer) should beabandoned, since they areinadequate to represent the diversityof behaviours of network users (whatkWh? what kW?).



Policy charges

In addition to network charges,in most power systems there is adiversity of other costs that have beentraditionally recovered from electricityconsumers (such as: subsidies todomestic fuels, clean technologies orsocial tariffs; charges for efficiency,energy conservation or innovationprogrammes; competition transitioncharges, etc.)

‘IN SOME COUNTRIES… THESE

ADDITIONAL REGULATED COSTS MAY

BE COMPARABLE TO THE AMOUNT OF

NETWORK COSTS.’

In some countries the total magnitudeof these additional regulated costsmay be comparable to the amount ofnetwork costs. Therefore the method ofallocation should not be taken lightly.Cost causality is weak or inexistent forthese costs, so it has to be decidedin the first place whether electricityconsumers should shoulder them ornot. Flawed allocation approachesmay encourage grid defection. It maybe advisable to include most of thesecosts in a general taxation system,unrelated to electricity rates.

Consumers should not be able toavoid them, as they are unrelatedto consumer or producer behaviour.However, this avoidance is possible ifpolicy charges are allocated using thesame criteria as for network charges.This is the case of ‘net metering’, whichis discussed later.

The value of aggregation

One can argue that, conceptually,under a well-designed system of pricesand charges, it should be immaterial ifDERs or any other agents, centralizedor decentralized, are or are notaggregated. What is, then, the value ofaggregation, and how does it dependon regulation?

Aggregation is the grouping of distinctagents in a power system to act as asingle entity when engaging in powersystem markets (both wholesale andretail) or when complying with powersystems regulations. Aggregators mayact as intermediaries between thecomplexity of the power system, withmultiple electricity services and a priceor charge for each one of them, andthe simpler signals that each agent iscapable of handling at a given instantin time.

Identifying the value of aggregation isrelevant, in order to determine whetherthis activity should be facilitated, oreven encouraged, or, on the contrarybe left to the initiative of the agents.Aggregation has many points incommon with retailing, and a keyregulatory issue is to ensure impartialprocedures for the access of allaggregators to trading platforms andagents’ data. Understanding the valueof aggregation is a prerequisite toaddressing the economic viability ofany proposed business model.

Three categories or sources of thevalue of aggregation can be identified:opportunistic, fundamental andtransitory. Opportunistic value emergesas a result of regulation or marketdesign flaws that allow an agent oraggregation of agents to increase theireconomic wellbeing without increasing– or even decreasing – the economicefficiency of the power system. This isthe case for the rules of allocation ofbalancing costs and procurement ofbalancing services in some countries.Netting peak consumption throughaggregation is another example. Socalled ‘net metering’ – the combinationof a single standard meter at theconnection point and mostly volumetrictariffs – results in a substantial subsidyfor the aggregation of demand andlocal generation, with the subsidizedamount being passed to the remainingnetwork users. The only solution to the‘net metering problem’ is to replace

the standard meter by an hourly oneand to follow the tariff design principlesexplained in the previous section.

Other cases of aggregation have valueunder present conditions, by increasingthe power system efficiency. Examplesinclude: sharing, among severalagents, some unavoidable costs ofparticipation in some electricity servicesmarkets; or facilitating the access ofsmall agents to hedging products; oreliminating the information gaps or thethreshold size barriers; or promotingthe technical capability to handle allthe relevant information associatedwith participation in complex markets.The fundamental component of thevalue remains, since it is intrinsic toaggregation and is independent of themarket or regulatory context. On theother hand, the transitory value mayextend some time into the future, butcertain technological or regulatoryimprovements will reduce its magnitudeuntil it disappears.

Conclusion

Any sensible long-term visions of thepower sector should be based onregulation that correctly implementsany future priorities of energy policy.Sound regulation will be essentialto efficiently blend centralized anddecentralized resources in futurepower systems. If regulatory innovationcannot keep pace with the changingnature of the electric power system,large inefficiencies may result. In theabsence of regulatory innovation,network users and new businesseswill find ways to arbitrage the growingdisconnect between ill-adapted

‘ANY SENSIBLE LONG-TERM VISIONS

OF THE POWER SECTOR SHOULD BE

BASED ON REGULATION THAT CORRECTLY

IMPLEMENTS ANY FUTURE PRIORITIES OF

ENERGY POLICY.’



regulations and new market and

technological realities.

The MIT Utility of the Future Study

is examining an ensemble of power

sector regulatory issues that need

a thorough review, and maybe also

significant adaptation, in the presence

of a substantial amount of DERs.

Wholesale markets may have to be

redesigned to include the active

participation of distributed generation,

demand, and storage. Regulators

need to reduce the uncertainty and

the information asymmetry in the

estimation of distribution network

costs, while incentivizing efficiency,but without jeopardizing innovation.The present organization of the powersector has to be reconsidered. Thisprocess will include issues such as: thecreation of neutral trading platforms,data ownership and management,the more complex new roles of theDistribution System Operators (DSOs),the relationship between DSOs andTransmission System Operators(TSOs), and the level of separationbetween commercial and regulatednetwork activities when conflicts ofinterest may exist. Finally, considerableattention should be paid to the

growing interactions that are alreadytaking place between the power andgas sectors, and the buildings andtransportation sectors, each with itsspecific regulations.

Regulation does matter. And most ofthe regulation that is mentioned herehas to be supported by quantitativeanalysis – such as a portfolio ofcomputer software tools that canrepresent the impact of DERs onthe electric power system underdifferent geographical and temporalperspectives. This will be the subjectof another paper.

Wind power and electricity supply security in ColombiaDavid Harbord, David Robinson, and Ivan M. Giraldo

The current regulatory regime inColombia is intended to be technologyneutral, but in reality it is not. This articledescribes some of the characteristicsof the Colombian electricity systemand how auctions are used to ensuresufficient ‘firm energy’ (the capacityto produce electricity reliably) to copewith uncertain hydrological conditions.It also explains why wind power wouldbe attractive for the Colombian systemand how current regulations discouragewind power investment.

The Colombian electricity system andauctions for firm energy

Colombia’s electricity system is heavilyreliant on hydro, which accountsfor about 64 per cent of its currentgeneration capacity and 80 per centof electricity produced during normalweather conditions. There is growingconcern, however, about how to ensuresupply security during periods of lowrainfall, specifically during El Niñoperiods such as occurred in 2009–10,when hydro’s share of electricitygeneration fell to 67 per cent (see the

table below). To avoid shortages, themissing hydro generation needs to bereplaced with generation from othersources.

‘THERE IS GROWING CONCERN

ABOUT HOW TO ENSURE SUPPLY

SECURITY DURING PERIODS OF LOW

RAINFALL…’

The traditional source of back-uphas been electricity generated fromthermal power plant burning coal ornatural gas. Many of these plants arenot required to generate electricity

during normal hydrological conditions.In 2010, which saw the effects of ElNiño at the beginning of the year,thermal plant accounted for 27 per centof electricity generation, while in 2011(not an El Niño year) this fell by 40 percent to 16 per cent of electricityproduced.

Because of their low load factors,many thermal plants would not recovertheir investment costs if they werebeing paid for solely on the basis ofrevenues earned in wholesale spot andcontract markets. In order to ensurethe building of sufficient generation

Generation output 2010–11 for the Colombian electricity system

2010 2011 ChangeGWH

Growth%GWH % GWH %

Hydro 38,088.6 67 45,583.1 78 7,494.5 19.7

Thermal 15,590.7 27 9,383.7 16 –6,207.0 –39.8

Minor plants 2,985.6 5 3,336.7 6 351 11.8

Cogeneration 222.7 1 316.9 1 94.1 42.3

Total 56,887.6 100 58,620.4 100 1,732.8 3

Source: XM, the company that operates and administers the Colombian wholesale electricity market.



capacity, in 2006 the Colombian energyregulator (the CREG) introduceda scheme to ensure the long-termreliability of the electricity supply, andin particular to guarantee the system’sability to meet peak demand during ElNiño periods. The scheme allocates‘firm energy obligations’ (OEFs) tonew and existing generation plantat prices determined in competitiveauctions. OEFs are ‘option contracts’that commit generating companies tosupply given amounts of energy at apredetermined Scarcity Price wheneverthe spot price in the electricity marketrises above that Scarcity Price. Thecompanies receive the spot price forany additional generation above theirfirm energy obligations, and pay apenalty if they cannot meet their firmenergy obligations.

In return for agreeing to supplyelectricity at the Scarcity Price,generators with OEFs receive a fixedannual option fee (the firm energyprice, or Cargo por Confiabilidad) foreach unit contracted. This option feemakes an important contribution to therecovery of fixed costs for generatingplants that produce very little in normaltimes – such as the CCGT plantsin central Colombia that generateinfrequently outside of El Niño periods.New plants chosen in the auction havea guaranteed revenue stream for 20years, even if they are never required togenerate.

The maximum amount of ‘firm energy’that a generator may offer in a firmenergy auction is known as its ENFICC(Energía Firme para el Cargo porConfiabilidad). A generator’s ENFICCrefers to the amount of energy of agiven type it can reliably and continually(in terms of kWh/day) produce duringperiods when hydro generation is ata minimum. It is essentially a lowerbound for the amount of energy thata given type of plant will be able toprovide at some unpredictable point inthe future. The following table shows

the typical ENFICCs for differentgeneration technologies in Colombiaas a percentage of a plant’s CEN(effective net capacity).

To date, there have been two firmenergy auctions (in 2009 and in 2011).In the first, about 9,300 GWh per yearwere allocated to new resources; thisincluded 1,117 GWh from new coalplant and 1,678 GWh from newgas-fired generation plant at anauction-determined option fee of13.998 USD/MWh. In the second,3,700 GWh of OEFs were allocated tofive new generation projects, with anoption fee of 15.7 USD/MWh. In thesecond auction, thermal plantaccounted for over 89 per cent of theOEFs assigned.

Firm energy from wind power in Colombia

There is growing interest in the potentialfor wind power to provide back-up forhydro, as an alternative to thermalgeneration plant. In part this is areflection of the falling cost of wind-based generation, but it also reflectsevidence that wind power potential isnegatively correlated with hydro incertain regions where wind speeds arehigher during dry periods. There arealso environmental reasons – inparticular lower carbon dioxideemissions – for preferring renewablesources of energy to hydrocarbons.

For wind power to be a viable

alternative to thermal generation, thefirm energy auctions need to accuratelyreflect the contribution of wind energyto system security. The challenge is todetermine the quantity of firm energythat can be provided by differentenergy sources when the system isunder stress, which in Colombiacorresponds especially to peak periodsunder El Niño (low hydro) conditions.For coal and natural gas, which can bestored, the auction regulations assumethat the plants can provide firm energyat about 90 per cent of their ratedcapacities. For wind power, on theother hand, the energy regulator(CREG) assumes low levels of firmenergy, about 6 per cent of ratedcapacity in 2011. Studies carried out bythe World Bank and by the OxfordInstitute for Energy Studies suggestthat the quantities of firm energy fromwind in some parts of Colombia aresignificantly higher, in the range of20–45 per cent.

The lower the firm energy attributed towind energy, the lower the revenuestreams that can be earned in firmenergy auctions. If the firm energycontribution of wind power has beenunderestimated, this acts as a barrierto investment in wind power inColombia, and indirectly raises the costof meeting system security. Thefinancial consequences for investorsare significant. Following the 2011CREG approach (at a 6.3 per cent firmenergy rating), a 100 MW wind plantwould earn approximately USD735,734per annum in guaranteed payments for20 years at the firm energy price(13.998 USD/MWh). Using an approachto measuring ENFICCs that wasproposed by the World Bank forColombia (which resulted in a 36 percent firm energy rating) a 100 MW windplant would earn approximately USD4.4million in annual firm energy payments.This makes a significant difference tothe financial viability of wind farms inColombia.

ENFICC % for differenttechnologies as of 2012

TechnologyMaximum

ENFICC (%)

Hydro with storage 55

Hydro without storage 30

Coal 97

Natural gas 93

Fuel oil 88



The question is whether the CREGreally does underestimate thecontribution of wind power to systemsecurity. To assess this, the authorslooked at the relevant economicprinciples and at selected internationalexperience.

Economic principles and internationalexperience

There is no universally acceptedmethod for calculating the contributionof intermittent generating technologies(such a wind) to system reliability.However, there are some basicprinciples that guide the methodologyto be used, as well as experiencein the application of thismethodology.

‘THERE IS NO UNIVERSALLY ACCEPTED

METHOD FOR CALCULATING THE

CONTRIBUTION OF INTERMITTENT

GENERATING TECHNOLOGIES TO SYSTEM

RELIABILITY.’

The main principle for calculating thecontribution of wind power to systemreliability is to reflect the amount offirm energy the system can rely onwhen there is a high risk of shortages.In most systems, this occurs duringperiods of peak demand. However, inColombia, the probability of shortageis highest during El Niño periods – inother words when hydro generation islow. So the question is how much firmenergy can be provided by wind powerin those periods.

One way of calculating a firm energyfactor is to use historic data todetermine the minimum amount ofenergy that can be provided by windpower in periods of system stress.Each system has different periods ofshortage, and each wind power stationwithin a system will have output thatcoincides, more or less, with thoseshortage periods. To the extent thatwind generation is higher at times of

shortage, the plant will have a higherfirm energy factor.

There are different ways to use the timeperiod-related data to approximatewind’s firm energy factor. One methodthat we think is sensible is used inPJM (the Pennsylvania–New Jersey–Maryland Interconnection, a regionaltransmission organization in the USA).This approach averages the wind-related generation over the relevantshortage periods in recent years inorder to estimate the level of firmenergy that can be expected during aperiod of system stress.

Application of the PJM methodology toColombia

The relevant shortage periods inColombia are mainly El Niño periods,especially during peak hours.Following the logic of the PJMapproach, we estimated the ENFICCsfor wind power using hourly generationdata from the experimental Jepírachiwind farm in Colombia’s Guajira region,between April 2004 and April 2011.The ENFICC estimate (shown inthe table below) uses the PJMmethodology, applied to wind outputon a daily basis and during peakhours during the last three El Niñoperiods. This yields average estimatesof ENFICC between 27 per cent and33 per cent, compared to the CREG’sestimates of below 15 per cent.

This suggests that the CREG’s original2011 methodology, with ENFICC below15 per cent is too conservative. Both weand the CREG measure ‘firm energy’. Intheir original methodology, the CREGmeasures it by reference to the lowestaverage monthly (in kWh-day) figure foroutput from the pilot wind plant; theirENFICC thus has a 100 per cent (or 95per cent) probability that wind outputwould be greater than the historicminimum. That plant-specific probabilityis too low because system reliabilitydepends on the probability of thecombination of plants providing energy,not simply on the probability related toan individual plant. This is especiallyimportant when there is an inversecorrelation between wind and hydrogeneration. Recently, the CREG haschanged its methodology and increasedthe ENFICC for wind. However, UPME, agovernment body responsible forlong-term planning, argues that themethodology is still very conservativeand ignores the complementaritybetween wind and hydro generation.

In contrast, we measure firm energy(using the PJM methodology) byreference to the historic average windoutput during the periods of significantsystem stress – which occur during ElNiño periods. By narrowing our focusto periods when the systems is understress, our ENFICC reflects wind’scontribution to system reliability whenhydro generation is low.

PJM methodology to determine ENFICC for wind power in Colombia

ENFICCs base 5–7 a.m.; 6–8 p.m. All day

Niño 1 24.13% 29.05%

Niño 2 27.32% 37.02%

Niño 3 30.34% 33.78%

Average 27.26% 33.29%

5–7 a.m.; 6–8 p.m. All day

All 3 Niños 27.52% 32.58%



Conclusion

At the beginning of 2014, wind power

in Colombia (all from the experimental

Jepírachi plant) accounted for only

0.1 per cent of total generation

capacity. There are new wind projects

under study, with the potential to

generate up to 674 MW (as of July

2015), but none have yet entered the

construction phase. However, the

current very high energy prices – which

are the result of reliance on diesel

oil-fired generation to replace hydro

power during the current El Niño event

– reinforce the need for a more

diversified energy mix, which would

include alternative sources like wind

power.

We think the reluctance to invest

is at least partly due to the current

regulations in Colombia, which

underestimate wind power’s potential

contribution to electricity system

security. Given the recent evidence

of the Colombian government’s

commitment to develop alternative

energies, particularly wind power, andits tradition of technological neutrality,the authors encourage a regulatoryrethink, in particular to recognize thecomplementarity among renewableresources in their contribution tosystem security. If this cannot bedone through the reliability auctionsdue to complexity or for other reasons,then other regulatory measuresshould be considered to takeaccount of this complementarity andto lower barriers to investment inwind power.

Decarbonization through electrification – the importance of energytaxation being in line with long-term energy policyGraham Weale

Electrification as a major instrument fordecarbonization

With the success of the COP21 Meetingin Paris the world has shown itself moredetermined than ever to combat climatechange. Europe has been at theforefront of this initiative since the late1990s and in 2010 the landmarkRoadmap 2050 was published by theEuropean Climate Foundation, showinghow the EU could achieve its 80 percent decarbonization aim. At the heart ofthe plan lay a greatly enhanced role forprogressively decarbonized electricitywithin the overall energy supply.Electricity would be needed to play aleading role in transport (supported bybiofuels and fuel cells) and also in thespace heating sector through theinstallation of heat pumps.

According to the Roadmap, electricitydemand would increase by up to40 per cent against the baseline by2050, with an additional 740 TWh and700 TWh being required respectivelyfor transport and heating (in both thehousehold and industry sectors).

This prospective growth of over1400 TWh should be put in the contextof final electricity demand in the EU-28of 2843 TWh (2010) and 2771 TWh(2013).

Whilst there are alternative solutions forthe future supply mix – whether thereshould be more emphasis onrenewables outside the power sector orwhether hydrogen may become animportant fuel for the transport sector– there is widespread agreement thatsubstantial electrification will be crucial.It then becomes important to seewhether the conditions are in place toallow this to be achieved and thenecessary infrastructure built upsmoothly along the way, rather thanbeing challenged by short- or medium-term declining electricity demand.

The current energy taxation policy ishindering electrification

The statistics for European electricitydemand in 2010 and 2013 cited abovepoint to an initial decline in demand ratherthan the start of a new electrification

trend. One factor contributing to this

development is that the energy carrier in

several countries (notably in Germany) is

loaded with considerably more tax and

other surcharges than competitive energy

carriers. Excise taxes on vehicle fuels are

renowned for being high, but (at least in

Germany) these have now been

overtaken by the taxes and surcharges on

electricity for electric vehicles.

Two important messages concerning the

market share of electricity in the final

energy consumption mix are seen in

graph overleaf. Most noticeable are the

very different positions which electricity

holds in different countries – 34 per cent

in Sweden and 20 per cent in the UK

– results which have not generally arisen

due to conscious planning on the part of

government. Only in France was a policy

introduced to place emphasis on the role

of nuclear electricity as a central means of

primary energy supply, in response to the

1973 oil supply crisis; this led naturally to

a relatively high electricity market share.

Sweden has a high electricity share due

to a very limited gas supply, abundant



hydropower, and a policy of not over-taxing this form of energy.

‘…ELECTRICITY’S MARKET SHARE IN

ALL FOUR OF THE LARGEST EU

ECONOMIES IS STARTING TO DECLINE.’

The second message from the graphis that electricity’s market share in allfour of the largest EU economies isstarting to decline. Of course therecould be various reasons for this, butthe classical price elasticity effect – dueentirely to taxation and surcharges – iscertainly one factor.

The graph on the right shows clearlyby how much electricity taxes andsurcharges have increased over thelast five years (far more than those ongas), principally as a result ofrenewables development in thesector. Eurostat has applied aparticular definition of taxes andsurcharges and therefore smalldifferences may be noted whencompared with other sources. Forexample, according to UK DECCStatistics in 2013 the total leviesamount to 0.026 EUR/kWh.)

The development of sound energypolicy is challenging at the best oftimes, and it is difficult to anticipateall the side effects every time. The

particular means of implementing and

financing national targets for

renewables in various EU countries is a

case in point, where the focus has

been on meeting renewables targets

and financing them by consumer levies.

It has proved practical to reach those

national targets for renewables by

concentrating on electricity, while some

countries had the additional aim of

replacing certain forms of thermal

power generation. As a result, in the

EU-28 over the period 2004–13 therenewables share of the power sectorincreased by 11.1 per cent; at the sametime, the increase in the heating sectorwas 6.6 per cent, while that in thetransport sector was 4.4 per cent.Both Germany and the UK saw twiceas much growth in the renewablesshare in their power sectors as intheir heating and transport sectorstogether. Conversely in Sweden muchmore emphasis has been given to theheating and transport sectors.

These developments in themselvesare not necessarily to be criticized,but the consequence which nowdeserves attention (and which wasmost likely not originally considered),is the relative increase in theend-consumer price of electricityas compared to competing sourcesof energy (mainly gas and oilproducts). Higher prices of energy ingeneral increase the incentive forenergy efficiency investments, anevidently positive outcome. However,high relative prices of electricity risksetting in motion a development thatmay threaten the long-term goal of

19%

21%

23%

25%

27%

29%

31%

33%

35%

2009 2010 2011 2012 2013

EU(28)

Germany

Spain

France

Italy

Sweden

UK

Market share of electricity in final energy consumption

00.020.040.060.080.1

0.120.140.160.180.2

EUR/kW

h

Gas2015

Gas2010

Elec2015

Elec2010

Taxes and surcharges on electricity and gas – household consumersNote: Values for 2015 are incremental over those for 2010. Denmark’s value for gas in 2015 was lower than its2010 value (0.483 EUR/kWh). Figure 2 shows the 2010 value.

Source: Eurostat – Prices and Taxes for Household Electricity and Gas Consumption.



decarbonization through electrification.In the worst case it could trigger theso-called ‘utility death-spiral’ (anexpression coined in the USA as thegrowth of consumer PV ate into utilityrevenue and profits), wherebyinfrastructure (required both to delivercentrally generated electricity and tosupport the growing decentralizedgeneration) would struggle to invest asrequired for the longer-termelectrification goals.

The problem of the rising taxes andsurcharges on renewables, shown inthe second graph, can becompounded by two further factorsthat can make the situation evenmore serious. Both relate todevelopments in the tariffs for thedistribution networks.

The first factor at play in somecountries is the cost of connectingrenewables, when this is borne bythe customers rather than the projectdevelopers and therefore translatesinto higher network charges. InGermany (including VAT), thesecharges have increased by up to0.02 EUR/kWh between 2010 and2015 to accommodate the increasednumber of windfarms, the productionof which is essentially exported tosouthern Germany, with the localconsumers enjoying absolutely nobenefits (other than modestemployment effects).

The second factor relates to thepernicious consequence of havingto spread a fixed cost operation(of which the power industry isbecoming an example parexcellence as the role of fossil fuelsdecreases) over progressively fewerunits of energy. There are twoelements at work here. First is theincrease in self-generation, reducingthe net consumption flowing fromcentral sources through the grid.The second is the decline in grosspower generation; this has been

catalysed by both the price elasticityeffect (an incentive to reduceelectricity consumption, togetherwith other energy efficiencyinitiatives), and the incentive tofavour, in relative terms, other formsof energy (the cross-price elasticityeffect), whether by switching awayfrom electricity or not switching to it(for example, heat pumpapplications).

The issue of the price elasticity effectscoming into play and working againstlonger-term electrification appears notyet to be showing up on governmentradar screens. If the current early trendscontinue, there is the risk that switchingaway from electricity may gathermomentum, leading to developmentswhich are in contradiction to the longer-term electrification aims, before suchswitching can be slowed down andreversed.

‘IF THE CURRENT EARLY TRENDS

CONTINUE, THERE IS THE RISK THAT

SWITCHING AWAY FROM ELECTRICITY

MAY GATHER MOMENTUM…’

Electrification, as one important meansof decarbonization, will most likely takeplace at the lowest cost if there is asmooth development from the currentposition out towards 2050. Therefore,the right long-term signals need tobe put in place to incentivize theprogressive development of electricityinfrastructure, rather than its prematuredownsizing.

Solution: bringing energy taxation policyinto line with energy and climate policy

How then can progress towardsmeeting EU and national renewablestargets be maintained and longer-termdecarbonization be kept on track?The answer lies in a root and branchreview of energy taxation policy, whichwould include renewables surcharges.Fortunately, even without designing

potential new systems there is a helpfulrange of policies (whether actual orunder discussion) from which to draw.

Option 1: Supporting renewables from

general taxation

Turning first across the Atlantic, thepolicy in the USA has been to supportrenewables mainly out of generaltaxation. Onshore wind farms currentlybenefit from production tax creditswhich are worth around 25 USD/MWh,and solar plants from 30 per centinvestment tax credits. The result is thatrenewables have grown withoutincreasing the consumer price ofelectricity and thereby reducing itsmarket share. Of course there havebeen other side effects of the policy.Long-term Power Purchase Agreements(PPAs) for wind power have beensigned at 25 USD/MWh or even lower,and these have artificially brought downthe wholesale price, which has putcertain thermal plants, in some caseseven nuclear, under pressure.

Option 2: Technologically neutral quota

systems

Within Europe, whilst feed-in tariffshave been the rule, some countries(such as Sweden and the UK) havefollowed quota systems. Swedenis one of the few countries to havefollowed a technologically neutralquota system. This led to a focus ona single technology at any given pointin time: first biomass (2000–7) andthen onshore wind. Other countries,in contrast, opted for the paralleldevelopment of more than one type oftechnology; this has the advantage thatsome of the more challenging formsof technology (such as offshore wind)which are essential to meet climategoals, have a greater chance of beingsuccessfully developed, and of beingbrought to commercial maturity morequickly, than under a technologicallyneutral support system. But it increasesthe costs over the short term.



Even though the quantities ofrenewable electricity which Swedenhas developed on the basis oftechnological neutrality are modest, itsaverage support costs are low, at only24 EUR/MWh. This compares with aweighted average of 95 EUR/MWh forthe eight EU countries with, in total,the most renewables, while the figurefor Italy – at 177 EUR/MWh – is almosttwice as high.

Option 3: Spreading the cost of

renewables across all forms of energy

France has an electricity surcharge:CSPE (Contribution au ServicePublic de l’Électricité). Approximatelytwo-thirds of this surcharge coversrenewables, the remainder goestowards other costs to be socialized(such as the higher costs ofelectricity in the French islands andcogeneration). The level of the CSPEwas 0.02 EUR/kWh in 2015 (less thanone third of the German equivalent) butthere is an ongoing discussion, partlybecause electricity prices are politicallyparticularly sensitive, as to whether thecost of the renewables subsidy shouldbe spread over more (or potentially all)forms of energy.

‘EVEN IN THE USA THE FUTURE OF

FURTHER TAXATION SUPPORT IS FAR

FROM CERTAIN…’

Whilst the public taxation option hassome compelling logic, the effortsbeing made by almost every nationalFinance Ministry to reduce their budgetdeficit makes it an unlikely runner. Evenin the USA the future of further taxationsupport is far from certain, and at thetime of writing the US legislators wereconsidering a plan to phase out all taxsupport over five years, with supportfor solar power potentially disappearingsooner.

The most realistic option then, onewhich should certainly be addressed,

is a careful reform of energy taxationwith the aim of ensuring maximumconsistency with long-term energy andclimate policy. There are a number ofdimensions to such a reform whichneed to be considered:

(a) The need to avoid disadvantagingelectricity by taxing itdisproportionately in relation toother fuels – the main theme ofthis paper.

(b) Ensuring, over the long run, that theprice of different forms of energy isclose to the full cost, includingexternalities.

(c) The need to ensure, above all else,a strong incentive for energyefficiency (but not for anyparticular fuel, other than asdictated by its overall carbonfootprint).

This would mean, first, that there shouldnot be fuel-specific taxes – such asan electricity tax rather than a gas tax.Also, to the extent that taxes as suchwould have revenue-raising objectives,then they should be proportional to theenergy element. (There is a differentargument in the case of motor fuelswhere taxes are needed to pay for theroad infrastructure.)

There may appear, at first sight, tobe a contradiction between (a) and(b) above, but renewables raise thechallenge that they originally neededsupport (and to some extent still do)in order to bring them to commercialmaturity. So far as the support requiredfor this purpose, as opposed to thefuture steady-state financing costs, canbe separated, the author is arguingthat such learning costs should eitherbe spread over all forms of energy or,where possible, covered from generaltaxation. In Germany, for example,almost half of the current annual0.063 EUR/kWh renewables supportcharge relates to the legacy supportfor PV. This cost component has norelationship to the long-term costs

of renewables electricity generationand including it as part of the finalconsumer electricity price (includingVAT, it represents around 12.5 percent of the price) is inconsistent withobjective (b) above.

The multiple political and environmentalrequirements placed on the energysystem, and in particular on thepower system, militate against aclean economic solution. In theprevious decade, the EU aimed,through market-based methods, tomeet the welfare objectives of theSingle Market (through the internalwholesale market – with progressiveimprovements to its working) andthe objective of decarbonization(through the carbon emission tradingsystem). However, these were quicklyand comprehensively usurped byvarious out-of-market mechanisms.It is now unrealistic to expect toomuch emphasis to be placed onthe (originally foreseen) simpleand fundamentally market-basedmethods to meet power demand anddecarbonize.

‘…THE NEED FOR ENERGY TAXATION TO

BE IN HARMONY WITH ENERGY POLICIES

AND STRATEGIES HAS SO FAR BEEN

COMPLETELY OVERLOOKED.’

However, whilst the wholesale marketshave been heavily distorted, leavingthem poorly equipped to achievemuch more than the optimaldispatching of existing plants, othereconomic forces are still very muchat work, including the aforementionedprice elasticity effects, which cannotbe ignored.

European governments are beingchallenged in many different directions,both by requirements to meet theirspecific 2020 goals, together withnational policy objectives. There isthus the risk that they give inadequateattention to the strategies required



to reach the 2050 end goal of 80–95per cent decarbonization (and at areasonable cost), and are pulled offcourse. The longer-term importance

of the need for energy taxation to bein harmony with energy policies andstrategies has so far been completelyoverlooked. It is now time to open

this book and make the necessarycorrections so that, over time, there willbe a smooth and optimal cost path tothe 2050 goals.

View from the USA: rapid transformation of power sector calls fornew tariffsFereidoon P. Sioshansi

Conventional electricity tariffs

It should come as no surprise toanyone that the cost of providingelectric service to most customers– notably residential consumers – ismostly fixed. Studies by the ElectricPower Research Institute (EPRI) andothers put the fixed component ofthe cost at around 60 per cent for thetypical residential user in the USA,which is the main focus of this article.What is surprising, however, is that forover a century, for most consumers,the bill has been mainly determinedby the volume of electricity consumed:a multiplier (cents/kWh) times thenumber of kWhs used. Even today,most residential consumers in the USApay virtually all their monthly bills onthe basis of volumetric consumption.Relatively little is collected through fixedfees or connection charges.

‘THE COST OF PROVIDING ELECTRIC

SERVICE TO MOST CUSTOMERS –

NOTABLY RESIDENTIAL CONSUMERS –

IS MOSTLY FIXED.’

The volumetric scheme did not makemuch sense when it started, but wasa convenient way to measure and billconsumers. For regulators, it was aneasy way to adjust the ‘multiplier’ everyso often to reflect changes in fuel,operating, or investment costs. Forconsumers, the concept made sense.Many still believe that electricity should

be billed on the basis of volume, aswith gasoline. If you don’t use much,you don’t pay much.

It did not much matter throughoutthe industry’s formative decades,when demand continued to grow,encouraging massive investmentsin generation and in transmissionand distribution infrastructure, whichcould be financed through the simplevolumetric scheme. With average retailtariffs flat or declining in real terms andrapid demand growth, everyone washappy – the consumers, the utilities,and the regulators.

Effects of distributed generation

Conditions are different today.Electricity demand in the USA – and innearly all other mature economies – isvirtually flat or in some cases actuallydeclining (see the table below). Asbuildings and appliances become

more efficient, less energy is needed tooperate them. Moreover, with the rapidfall of the cost of distributed generation(DG), notably in rooftop solarphotovoltaics (PVs), consumers canproduce more of what they consume,which means less is bought from thenetwork, the grid.

While the impact of DG is uneven, insome places its effects are becomingpronounced. For example, Energex,a distribution utility serving theBrisbane metropolitan area in sunnyQueensland, Australia already hasnearly 300,000 customers with rooftopsolar PVs, making it among the higheston a per capita basis.

Making matters worse is the massivecost of maintaining and upgradingan ageing infrastructure upstream ofthe meter – the costs of which mustbe spread across a shrinking base.This leads to higher retail tariffs, which

Get used to it: no growth in utility business – New York or elsewhereAverage electric sales growth in New York, 1966 to 2013 and projections to 2024

Period Avg. sales growth for period

1966–76 3.8%

1976–86 1.5%

1986–96 1.4%

1996–2006 0.9%

2003–2013 0.3%

2014–2024 0.16%Source: New York Public Service Commission, 26 February 2015.



encourages more investment in energy

efficiency and DG, which leads to

even higher tariffs – the so-called utility

death spiral.

‘ELECTRICITY TARIFFS ARE THUS IN NEED

OF OVERHAUL AS CONSUMERS BUY LESS

AND GENERATE MORE.’

Not surprisingly, consumers in many

European countries, where electricity

is heavily taxed and/or loaded with

levies, are also looking for any

measure at their disposal which would

enable them to buy less; this means

using less and self-generating more

when it is cost-effective to do so.

Electricity tariffs are thus in need of

overhaul as consumers buy less and

generate more.

Future of electricity tariffs

In this context, the debate on whatto do with electricity tariffs is heatingup in many parts of the world, asregulators and politicians grapple withfinding ways to keep the incumbentssolvent without necessarily crushingthe rapid uptake of solar PVs, whichare enormously popular amongconsumers.

A recent article in The Wall StreetJournal (20 October 2015), entitled‘As Conservation Cuts ElectricityUse, Utilities Turn to Fees’, describedthe dilemma faced by utilities andregulators in the USA as they try toadjust consumer tariffs by shifting moreof the costs of service to fixed fees –which is how it should have been in thefirst place. It said:

Electric utilities across the countryare trying to change the waythey charge customers, shiftingmore of their fixed costs tomonthly fees, raising the hacklesof consumer watchdogs andconservation advocates.

Traditionally, charges forgenerating, transporting andmaintaining the grid have beenwrapped together into a monthlycost based on the amount ofelectricity consumers use eachmonth. Some utilities alsocharge a basic service fee of$5 or so a month to cover thecosts of reading meters andsending out bills.

Now, many utility companies areseeking to increase their monthlyfees by double-digit percentages,raising them to $25 or more amonth regardless of the amountof power consumers use. Theutilities argue that the feesshould cover a biggerproportion of the fixed costsof the electric grid, includingmaintenance and repairs.

The basic logic of what the utilitiesare trying to do makes sense, butchanging utility tariff structures iscomplicated, slow, and convoluted.Making matters worse, it is highlypolitical. Any change in current tariffsmeans shifting costs to othercustomers, who do not like payingmore.

Many believe that the electricity grid isessentially a public good and,therefore, must be paid for by everyonewho benefits from it.

Utilities in at least 24 states haverequested higher fees. Confronted withthe facts, regulators in a number ofstates are sympathetic to the plight ofutilities but are reluctant to raise fixedfees too aggressively.

Fixed fees have their critics. Consumeradvocates point out that theydisempower the customer anddiscourage investments in rooftop solarand energy efficiency.

Fundamentally, the thinking is tomove towards tariffs that reflect

cost-causality so consumers getcharged for the costs they impose onthe network while raising sufficientrevenues for the maintenance andupgrading of the grid and thereliability that it offers. These includeseveral well-known tariff optionssuch as:

Time-of-use (TOU) rates;

Increased fixed charges;

Three-part tariffs; and

Demand subscription.

Three-part tariffs are getting sometraction as they purport to closelyfollow the costs attributed to servingcustomers in the emerging businessenvironment. One such tariff introducedby the Salt River Project (SRP), a non-regulated utility operating in Arizona, forexample, consists of the following threecomponents:

A fixed charge of USD18–20regardless of usage level, peakdemand, or anything else. It may beconsidered a connection or networkfee;

An energy charge, slightly lower thanthe previous volumetric charge, torecover the variable cost of fuel and/or purchased power; and

A demand charge, which variesbased on how customers’ peakdemand coincides with the network’speak demand.

For a customer with an 8.5 kW solarpanel, for example, the demandcharge varies from USD41/month inthe winter to USD126/month in thesummer.

‘THE REGULATORS DO NOT WISH TO

ANGER THE SOLAR CUSTOMERS BUT

ARE BECOMING SENSITIVE TO THE

PLIGHT OF THE UTILITIES…’

Other states examining three-parttariffs include Nevada. Describing the



merits of its recent application for solarcustomers filed with the Public UtilitiesCommission of Nevada (PUCN) inearly August 2015, Kevin Geraghty,NV Energy’s vice president of EnergySupply said:

A three-part rate design betterreflects the unique costs ofserving our net meteringcustomers and eliminates theunreasonable shifting of costsbetween those that can accessrooftop solar and net meteringand those that don’t, [addingthat] This is a proven ratestructure that has been in useby our commercial customersfor more than 50 years. Underthe proposal, net meteringcustomers still have theopportunity to reduce their billfrom NV Energy if they reduce theimpact they have on the grid.

The regulators do not wish to anger

the solar customers but are becomingsensitive to the plight of the utilitieswhich are suffering from revenue

erosion while higher costs are beingshifted to non-solar customers. Thebattle is just beginning.

A comparison of US and EU electricity prices: the relevance of thegovernment wedgeDavid Robinson

This article is a short English summaryof my recent report, published inSpanish (‘Análisis comparativo de losprecios de la electricidad en la Unión

Europea y en Estados Unidos: Unaperspectiva española’), comparingtrends in final electricity prices in the EUand the USA over the period 2008–14.

The article explains why final consumerelectricity prices in the EU, especially inSpain, increased significantly more thanin the USA between 2008 and 2014.

168

47

44

25

24

23

18

13

12

11

0 50 100 150 200

Hawaii

California

Arizona

NewJersey

Colorado

Vermont

Massachusetts

Louisiana

Connecticut

Maryland

Wattspercapita

Sunshine states: states that get most electricity per capita fromresidential solar panels, 2014Source: Rebecca Smith and Lynn Cook, ‘Hawaii wrestles with vagaries of solar power’, The Wall Street Journal, 29June 2015.

90

100

110

120

130

140 Residential-EurostatDD

90

100

110

120

130

140 Commercial -EurostatIBEU28 USA

90

100

110

120

130

140 Industrial -EurostatID

1. Evolution of average final electricity prices in the EU and the USA, 2008–14Source: Eurostat and US DOE-EIA



It concludes that the primaryexplanation is not related todifferences in conventionaldeterminants of electricity price (thedemand for and the cost of wholesaleelectricity and networks) but rather todifferences in taxes, levies, and othercharges related to financing publicpolicies, notably support forrenewable power.

Comparing final prices

Graphic 1 includes electricity price

indices for residential, commercial

and industrial consumers in the two

regions since 2008. They illustrate

clearly that average prices in the EU

have risen significantly faster than in

the USA. For instance, prices for

residential consumers rose by 34

per cent in the EU (EU 28) and by

18 per cent in the USA. For industrial

consumers, average electricity prices

rose by 22 per cent in the EU and by

6 per cent in the USA.

‘…AVERAGE PRICES IN THE EU HAVE

RISEN SIGNIFICANTLY FASTER THAN IN

THE USA.’

At the beginning of the period, USprices were much lower than in the EU.

By the end of the period, the difference

was even greater, with average EU

prices approximately twice as high as

in the USA for each of these customer

categories (see Graphic 2).

Why is there a large divergence in final

price trends?

I consider various possible

explanations for divergent price trends,

including differences in electricity

demand and in the two most important

costs of electricity supply: wholesale

electricity market prices and the cost of

networks.

Demand

The average demand per household

in the USA is more than twice the

average level in the EU (4,137 kWh in

2009). Since the electricity industry has

high fixed costs, higher volumes for

any given fixed costs usually translate

into lower unit costs and prices, as

illustrated by Graphic 3, which plots

prices and consumption per household

in the US states.

Although lower EU consumption per

household may help to explain higher

average prices at any given time, it

does not explain why average EU

prices have risen so much faster than

0.00

0.05

0.10

0.15

0.20

0.25

ResidentialEurostatDD

CommercialEurostatIB

IndustrialEurostatID

€/kW

h

EU28 US

2. Final electricity prices in the EU and the USA, 2014-S2Source: Eurostat and US DOE-EIA

0

50

100

150

200

250

300

350

400

6,000 8,000 10,000 12,000 14,000 16,000

Average

price($/M

Wh)

Averageannualconsumption(kWh)

3. Price per MWh versus average annual electricity consumption per household in US states



in the USA since 2008. If US electricitydemand had grown significantly fasterthan in the EU, this might help toexplain the faster rise in EU averageprices. In fact, US electricity demandwas approximately the same in 2014and in 2008, whereas EU electricitydemand fell by about 3 per cent overthe period. That difference may explaina small part of the price divergence,but not very much. For instance, if thefixed costs of networks were 30 percent of the final price, a difference of3 per cent in demand growth wouldtranslate into an increase of about1 per cent in the final price.

Wholesale market prices

The availability of low-cost shale gashelps to explain why US wholesaleelectricity prices have not risensince 2008 – indeed, in nearly all USwholesale markets, they have fallen, asillustrated in the right side of Graphic4. However, this does not explain thedifference in final consumer pricetrends because wholesale electricityprices follow a similar pattern of flat orfalling prices, depending on the dateschosen, in both the USA and the EU.In the EU, prices have been depressedby a combination of excess generation

capacity, low coal and carbon dioxideemission allowance prices, depresseddemand, and high penetration ofintermittent renewable electricity.

Network costs

The costs of transmission anddistribution networks have increasedin the EU and the USA at roughly thesame rates over the period, so they donot explain the diverging price trends.

I conclude that the diverging paths offinal prices between the USA and theEU cannot be explained convincinglyby any of the traditional demand orsupply cost factors, alone or together.

The government wedge

The term ‘government wedge’ refers totaxes, levies, and other charges thatfinance public policies through theelectricity tariff. The wedge does notinclude the cost of policies that arefinanced through the governmentbudget; it only covers those costs thatare financed through tariffs. The wedgeis the third component of electricityprices, with the other two beingcompetitive energy prices (forwholesale energy and retail services)and regulated network costs.

The government wedge has beengrowing steadily in most Europeancountries since about 2008. Accordingto the electricity industry lobbyEurelectric, between 2008 and 2012 thegovernment wedge rose by 31 per centfor residential consumers and by over100 per cent for industries, whereaswholesale energy costs fell and networkcosts rose by only 10 per cent and17 per cent for residential andindustrial consumers, respectively.

To take one example: in Spain,final electricity prices for residentialconsumers have risen by over 50per cent since the end of 2008. Thegovernment wedge explains over 70per cent of the increase and in 2014it accounted for 46 per cent of thefinal price for these customers. VATand special electricity taxes are themost important component of thewedge for residential consumers.Then comes the financial support forrenewable electricity and cogeneration,and interest payments on the EUR25

0

20

40

60

80

100

120

140

Spain Italy

Germany France

0

20

40

60

80

100

120

140

PJM(load-weightedaverage)

ISO-NE(WCMassachusetts)

CALIFORNIA(SP15-SOUTH)

ERCOT(LZNorth-Texas)

4. Selected EU and US wholesale electricity market price indices, 2008–14Sources: OMIE, EEX, GME, ERCOT, PJM, CAISO, ISO-NE, MorningStar

‘THE TERM “GOVERNMENT WEDGE”

REFERS TO TAXES, LEVIES, AND OTHER

CHARGES THAT FINANCE PUBLIC

POLICIES THROUGH THE TARIFF.’



billion tariff deficit (in other words,the accumulated difference betweenregulatory entitlements and thetariffs set to collect the cost of theseentitlements). The wedge also includes,inter alia, funding for domestic coalsuppliers, compensation for highercosts of supply on the islands, andpayments to certain consumers.

The government wedge in the USA isdifficult to measure, largely becausepublished data for the sector bundlesany policy-related costs into thecategories of generation, transmission,and distribution. It is also difficult tocompare states which have verticallyintegrated and regulated electricitymonopolies with states that haveunbundled the sector and havecompetitive wholesale or retail markets.

Nevertheless, the available evidencesuggests that the wedge is muchsmaller in the USA than in the EU. Thefirst piece of evidence is the greaterinvestment in, and penetration of,intermittent renewable energy in the EU,as illustrated in Graphic 5. On average,intermittent renewables account formore than 6 per cent of EU electricitygeneration and for less than half thatpercentage figure in the USA.

Second, the USA finances anincreasingly significant part of thecosts of renewable power and other

public policies (such as energyefficiency and smart grids) throughfederal and state taxes, rather thanthrough electricity tariffs. Total federalsubsidies for renewable power – bothdirect payments and tax credits – wereover US$11.7 billion in 2013, up fromabout US$1 billion in 2007. The federalgovernment also provided subsidiesof US$1.2 billion for smart grids andtransport in 2013. The US Congresshas recently extended the regime ofinvestment credits for solar energy andproduction credits for wind power.

In addition to federal subsidies for thepower sector, there are many state-levelsubsidies. Texas, for instance, offers taxcredits for renewable power that wereworth approximately US$1.4 billion in2013, up from about US$700 million in2007.

Third, most European countries levyVAT and other taxes on electricityconsumption, whereas in the USAelectricity is often not taxed, or is taxedat a much lower level. The averageVAT levied on electricity in the EU rosefrom 15 per cent in 2008 to 17 percent in 2012. This is not an issue forcommercial and industrial consumerssince they can recover these taxes.However, it does have an importantimpact on residential consumers whoare unable to recover these taxes.

Implications and recommendations

The immediate consequence of risingprices in the EU has been to reducethe disposable income of residentialconsumers, especially in countries witha very high penetration of renewablepower, notably Germany and Spain.Although industry and commercehave also seen rising prices, the bulkof the extra costs have been born byresidential consumers. This is the casein Spain, as illustrated by Graphic6, where the top part of the graphicreflects the government wedge, withthe residential consumers on the left,commerce in the middle, and industryon the right hand side.

A second implication is the distortingeffect on energy markets. Facing finalprices that exceed the (long or shortrun) marginal costs of electricity supply,consumers are encouraged to behavein ways that are not efficient, inparticular to discourage the use ofelectricity and instead to encourageconsumption of fossil fuels, for instancein heating. If the EU is to achieve itspolitical objective of reducinggreenhouse gas emissions by 80–95per cent by 2050, this will requirealmost complete decarbonization of theelectricity sector and large-scaleelectrification of transport, heating, andeven industrial energy consumption.

0

50

100

150

200

250

300

350TWh EU28

0

50

100

150

200

250

300

350TWh US

SolarWind

5. Electricity generation from intermittent renewables in the EU and the USASource: Eurostat and US DOE-EIA



I make a number of policyrecommendations. The first is that EUcountries should be rethinking how bestto finance public policies, especiallyin view of the Paris Agreement. It isevident that in most countries, electricitydecarbonization is a necessary andefficient way of reducing carbon dioxideemissions. This will involve additionalcosts for society, at least initially. I thinkit is a mistake to finance the additionalcosts of decarbonization incurred in theelectricity sector almost entirely throughthe electricity tariff, especially given thedistributional consequences.

Instead, I favour the introduction of arising minimum carbon tax on fossil fuelsspecifically for those activities (such astransport and heating) not covered by theEuropean Union Emission TradingScheme. This tax should also help to levelthe competitive playing field betweenelectricity and fossil fuels. The carbon taxshould be neutral from a fiscal perspective,in the sense that there should becorresponding reductions in other taxes,with the aim of having a positive netimpact on the economy (in terms of, forexample, GDP, public debt, employment).

Second, for public policies that havenothing to do with the efficient and securesupply of low-carbon electricity, I favour

recovering some of the related coststhrough the general government budgetrather than through electricity tariffs.This would have positive distributionalconsequences since income tax isprogressive whereas high electricityprices are a burden on the poorestconsumers. In principle, it also makessense for public policies that aim to meetbroad European or Spanish politicalobjectives (industrial, regional, or social)to be financed through the tax systemrather than through electricity tariffs.Nevertheless, along with the carbon tax,this sort of fiscal reform requires carefulstudy to ensure that the outcome isoptimal for the economy as whole.

‘TRANSPARENCY IS REQUIRED NOT ONLY

FOR FAIRNESS, BUT ALSO TO ENSURE A

LEVEL PLAYING FIELD IN THE EU…’

Third, in many EU countries, we needgreater transparency about whichcosts are being collected throughthe electricity tariffs, who is receivingthe benefits (subsidies), and who isbearing the burden. In some countries,notably Germany, there is significanttransparency. In others, notably Spain,there is very little. Transparency is requirednot only for fairness, but also to ensure a

level playing field in the EU when it comesto knowing whether state aid is beinggranted to certain industrial consumers.

Fourth, with respect to tariffs, thechallenge is not simply to reduce thelevel of final prices, but to design anefficient network tariff structure to recoverrecognized costs. Redesigning tariffs tocollect fixed costs through a fixed chargemay help to overcome the incentiveto bypass the system. However, thehigher the government wedge includedin the tariff (whether in the fixed orvariable component), the harder it willbe to sustain these tariffs, politically andeconomically, especially as the costs ofdistributed generation fall.

Finally, whether public policy costs ofdecarbonization are financed through theelectricity tariff or through the budget, weare left with the same underlying problem:a growing tension between existingelectricity market design and the need todecarbonize the economy. To date, thishas resulted in growing governmentintervention that has distorted andpossibly broken electricity markets. It istime to rethink the design of electricitymarkets and their regulation, to beconsistent with decarbonization and withincreasing decentralization of energyresources.

0.00

0.05

0.10

0.15

0.20

0.25€/kWh Residential-EurostatDC

Commercial-EurostatIC Industrial-EurostatIF

GovernmentwedgeNetworksEnergyandsupply

6. Composition of average final electricity prices in Spain for residential, commercial, and industrialconsumers in 2013-S2Source: Eurostat and own calculations



Introduction

Interest in demand response (DR– sometimes referred to under themore general heading of demandside management, DSM) has beengrowing worldwide and there are somefamiliar examples in the OECD. Forinstance, the USA has recently seena considerable growth in demandresponse aggregators such asENERNOC, which manages around25 GW of peak load DR capacity,and the PJM (Pennsylvania–NewJersey–Maryland) interconnectionin the north-east, which has welldeveloped DR markets. In the UK,demand response has, for some time,been an element of the National Grid’sShort Term Operating Reserve andthe Grid has recently been developingnew instruments like the DemandSide Balancing Reserve. The ideaunderlying all these approaches is thatit is often better to deal with periods ofimbalance between electricity supplyand demand by adjusting demandrather than generation. Supply anddemand have to be kept in balanceat all times because of the difficulty ofstoring electricity; traditionally mostsystems have relied almost exclusivelyon flexibility on the supply side (indeedmany still do). Including demand inthe calculations may not only be moreefficient, it may also help to expand thewhole debate.

‘[CHINA] IS STARTING TO EXPERIMENT

WITH A MORE ECONOMIC APPROACH TO

DEMAND RESPONSE.’

While developments in the OECD arefamiliar, it should not be forgotten thatthere are also examples from outsidethe OECD, in particular in China,which is the focus of this article.

China has, in fact, practised a form ofadministrative demand managementfor some time. However, as it movestowards the greater use of priceincentives within its electricity system,it is starting to experiment with a moreeconomic approach to demandresponse. In 2014, Shanghai wasselected by the National Developmentand Reform Commission (NDRC) asthe first pilot city in China to trialmunicipality-led DR programmes.The NDRC has also designated fourother municipalities (Beijing, Suzhou,Tangshan, and Foshan) as DSM citiesand instructed them to undertakeDR pilots. As part of this process, theOIES, working with the EnvironmentalChange Institute (ECI) of OxfordUniversity, undertook an assessmentof the Shanghai pilot on behalf of theNatural Resources Defense Council(NRDC – not to be confused with theNDRC!). Their work (available on theNRDC website as: ‘Assessment ofDemand Response Market Potentialand Benefits in Shanghai’, datedSeptember 2015, and referred to inthis article as ‘the Report’) is designedto provide an assessment of thepotential for, and benefits of, demandresponse in Shanghai, drawing bothon the outcome of the pilot and onexperience across the OECD. Thisarticle summarizes the findings ofthe Report.

Chinese context

In many ways, the factors drivingChina’s interest in DR are the same aselsewhere in the world, and reflect anumber of important trends, includingthe following:

Increasing penetration ofintermittent generation and, as aresult, declining flexibility on the

supply side. China, like many othercountries, has been investing heavilyin renewable sources and inparticular in wind power – windcapacity has increased more than300-fold since the beginning of thiscentury. The decreasing controllabilityof the supply side, resulting from theintroduction of intermittent sources,makes flexibility on the demand sidemore valuable.

Growth in demand side flexibility.With the development of smart gridsand smart meters (and localizedgeneration, which often actseffectively as a demand-sideresource) the cost and practicality ofproviding flexibility on the demandside is growing, and it is becomingeasier to coordinate large numbersof distributed loads.

Liberalization of electricitymarkets is also helping toencourage more flexible andinnovative approaches to pricingand system management – systemoperators are becoming moreopen to the idea that not everythinghas to be controlled from thecentre.

Finally, government policy is alsoincreasingly favourable to the idea ofdemand response. Concerns aboutclimate change are leading toincreased emphasis on reducingcarbon emissions and encouraginglower resource use – the powerstation which remains unbuilt, andthe tonne of carbon not emitted, areseen as valuable goals inthemselves.

However, it is also important to takeaccount of the very different design andhistory of the Chinese electricitysystem. In particular, it has in the pastput much less emphasis on economic

Assessment of demand response in ShanghaiMalcolm Keay, Xin Li, and David Robinson



incentives and price signals than isseen in the OECD. Consumer pricesare regulated and do not necessarilyreflect the full cost of supply. Systemoperation is also generally on anadministrative basis rather than on thesort of merit order we are familiar within the West – dispatch of powerstations takes place according togeneration plans which guarantee acertain level of output for generators,as a way of offering them a secureincome stream. Interregional andinterprovincial power exchanges alsonormally take place according to afixed schedule. To balance supply anddemand, utilities in China havetherefore operated administrative loadmanagement programmes, includingload shifting and avoidance, andrestrictions on electricity use orcurtailments in the event of severepower shortages. As a result of theseprogrammes, and of the importance ofindustry in Chinese power demand, theprofile of electricity demand in China isvery flat by OECD standards.(Shanghai is not entirely typical of theChinese system; it has a largecommercial and residential load, andincreasing air conditioning demand; itoperates a form of merit order basedon generation efficiencies; and it alsoimports a significant proportion of itspower from outside.)

‘TO BALANCE SUPPLY AND DEMAND,

UTILITIES IN CHINA HAVE THEREFORE

OPERATED ADMINISTRATIVE LOAD

MANAGEMENT PROGRAMMES…’

However, the Chinese electricity systemis changing. In particular, the authoritiesare encouraging the use of moremarket-based incentives, and the DRexperiments are part of this initiative(see ‘Opinion on further Deepening thePower Sector Reform’, a reform planissued by the State Council in March2015). In accordance with this process,voluntary interruptible contracts

between customers and utilities arebeing developed. Electricity demand isalso changing as the balance of theeconomy shifts from industrialproduction to a greater share forconsumption and public services.Also, given the growth of the airconditioning load in cities such asShanghai, the ‘peakiness’ of electricitydemand is likely to increase; the DRpilot is aimed, inter alia, at reducing theneed for new gas peaking plants.Finally, China is, like other countries,increasing its focus on environmentalissues: from local pollution (forinstance, coal-fired power plants,other than combined heat and power,are not allowed in the eastern coastalregions such as Shanghai) to globalclimate change (China has a targetthat carbon dioxide emissions shouldpeak by 2030 or earlier). The Shanghaipilot was set up against thisbackground.

Potential

A number of factors have to be takeninto account in calculating the potentialof DR in Shanghai. In particular,the Report considered two mainbenchmarks:

Experience with the 2014 pilot.The pilot was relatively limited inscale. It involved 27 large commercialcustomers who were asked to reducedemand during peak periods (whichgenerally take place between 1 p.m.and 3 p.m. on summer days,because of the air conditioning load).It achieved an overall reduction of 9per cent in demand among thesecustomers, although individualcustomers managed to reduce theirloads by much more significantamounts, while higher overallreductions (up to 30 per cent) werealso achieved over short periods (upto 15 minutes). Amongst otherthings, the pilot was looking at thepotential for developing the role of

aggregators (on the same generallines as in the USA).

International experience.Experience in other countriesprovides a general guide to thepotential, but clearly it cannot betranslated directly. Much dependson the policy instruments used(such as direct control of loads vsprice signals); the structure ofsupply and demand in the systemsconcerned; the nature of the marketinvolved; and the precise form ofdemand response (such as whetherthe service being provided isshort-term frequency control or awider contribution to reducing peakcapacity).

Taking account of these factors, andof the policy instruments which mightbe used in Shanghai, calculationswere made of the potential reduction inpeak capacity in the period up to 2030under various scenarios. On a relativelyoptimistic scenario, the analysisshowed that the market potential of DRresources could reach 2.5 GW in 2030,or about 4 per cent of peak demandin that year, though more conservativescenarios were also presented underwhich the reductions could be less– as little as 214 MW on the mostpessimistic forecast.

Benefits

The benefits of DR extend throughoutthe electricity system – in principle, itshould reduce the need for newgeneration capacity and, potentially,for new transmission and distributioncapacity. It should also reduce theneed for generation from existingfacilities, together with the emissionsentailed. In addition, it may contributeto a reduction in system balancingcosts and provide services such asfrequency response. Most of thesebenefits come in the form of avoidedcosts – that is, they cannot beobserved directly but must be



estimated. The costs of demandresponse are also system andcustomer specific and, especiallygiven the need to look into the future,not calculable with completeprecision. Indeed, the Report focusedon avoided costs and did not assessthe costs of demand response.Nonetheless, it attempted to producean indicative overall assessment of thebenefits.

‘[DR] SHOULD REDUCE THE NEED FOR

NEW GENERATION CAPACITY AND,

POTENTIALLY, FOR NEW TRANSMISSION

AND DISTRIBUTION CAPACITY.’

The biggest single element is theavoided generation capacity cost(taken as gas-fired plant, as requiredby regulation in Shanghai). Amethodology often used in the OECDto calculate the avoided capacity costis known as netCONE (CONE stands

for cost of new entry; netCONE isbased on annualizing the capital andother upfront costs of a referenceplant, taken to be representative of newplants on the system, and then takingaccount of expected market revenuesto produce a net figure). The aim is tocalculate the annual capital costs ofnew plants, which might be avoidedby DR programmes. There are manyestimates of such costs from OECDsources – for instance, in the UK afigure of 49 GBP/kW for netCONE wascalculated as part of the preparationfor the capacity auctions (though inpractice the outcome was much lower,at less than 20 GBP/kW, partly becauseof the number of existing plants whichwere successful in the auction, partlybecause of the number of diesel plantsinvolved – the reference plant for theCONE calculations was gas-fired).

As these complications suggest,such calculations involve a number

of assumptions and can never bedefinitive; furthermore, it is not easyto translate the assumptions orcalculations directly from one system toanother. In addition, although there aremany estimates of construction costsfrom OECD sources, construction costsfor new capacity in China tend to bemuch lower, so in the ECI/OIES study,estimates of Chinese costs were usedrather than OECD ones. The value ofthe potential savings also depends toa significant extent on the discount rate(for which various alternatives wereused for the purposes of the study)though discounting is not commonpractice in China. The study thereforeproduced a range of possible figuresfor the savings; benefits from capacityavoidance of around RMB550 millionin 2030 were possible on the moreoptimistic scenarios.

The study also estimated other avoidedcosts, including those of energy,

Total avoided costs



carbon dioxide, and transmissionand distribution, though these wereof a lesser order of magnitude. Totalavoided costs were estimated at alittle over RMB800 million in 2030at a discount rate of 7 per cent onthe most optimistic scenario, withcorrespondingly lower figures for thepessimistic scenario (see the figure‘Total avoided costs’). However thestudy also warned that the figureswere only indicative and might in somerespects overstate the value of DR –

in particular because of the practiceof administrative demand planning,which might have suppressed someeconomic demand, leading to a lowerlevel of capacity than is economicallyjustified, and because the studywas also unable to include the costsinvolved in DR.

Conclusions

Despite the various caveats notedabove, the study was particularly useful

as a ‘proof of concept’ in showing thatDR could, in principle, be a usefulpolicy tool in China, as in the OECD.It included recommendations for thedesign of DR policy instruments inChina and for further research todevelop a more robust evidence basefor calculation of the potential of DR.Some of this work is now beingundertaken and it is likely that, aselsewhere in the world, DR willincreasingly be an important part ofthe electricity scene in China.

Power system reform to enable large-scale wind utilization in ChinaZhang Xiliang and Xiong Weiming

Wind development and curtailment gohand in hand

Recognizing that wind power is oneof the most mature, abundant, andrich sources of renewable energy,the Chinese government has madeconsiderable efforts to promote it forthe mitigation of carbon emissions andimprovement of environmental quality.Over the last decade, China’s windindustry has enjoyed dramatic growth,overtaking the USA as the biggest windpower market in 2010. By the end of2014, China’s total wind power capacityhad reached 96.4 GW, accountingfor 7 per cent of the country’s totalgeneration capacity.

‘…CHINA’S WIND INDUSTRY HAS

ENJOYED DRAMATIC GROWTH,

OVERTAKING THE USA AS THE BIGGEST

WIND POWER MARKET IN 2010.’

Despite this successful increase inwind capacity, however, since 2010the industry has faced new challengesdue to its ineffective utilization. In thefirst half of 2015, wind curtailmentwas as high as 17.5 TWh at nationallevel, meaning that 15.2 per cent of

technically available wind power couldnot be connected to the electricitygrid, so the wind turbines had to shutdown. The problem of wind curtailmentis more prominent in the wind-richprovinces. In Jilin, Gansu, and Xinjiang,the proportion of curtailed wind powergeneration in available wind outputhas increased to 43 per cent, 31 percent, and 28 per cent respectively,making a negative impact on ongoingwind development. Although thegrowth of wind installation is still onthe fast track, the ineffective utilizationof wind capacity has caused energywaste, extra carbon emissions, andlower incomes for wind farms; thiscontradicts the original intention ofensuring significant priority to thepromotion of wind power in nationalenergy strategies. Hence, it is crucialfor the future development of windpower in China to overcome windcurtailment issues.

What led to wind curtailment?

In order to address this issue, a deepanalysis – to identify the obstaclesconstraining the large-scale utilizationof wind power – is necessary. From

the technical perspective, curtailmentis caused by the intermittency andvolatility of wind power itself. However,we argue that it is essentially due tothree main factors:

1 uncoordinated planning betweenwind and grid,

2 inflexible dispatch mode, and

3 the lack of policy incentives forstakeholders.

Firstly, inconsistency between winddeployment and grid planning haslimited the transmission of wind power.China’s rich wind energy resourceis mainly concentrated in the ThreeNorth area (north-east, north-west,and North China), which accounts forapproximately 90 per cent of nationalwind potential. In order to achievelarge-scale development of the mostwind-rich areas, central governmenthas set the ambitious target ofachieving seven large-scale (10 GW)wind farms by 2020, with six of thembeing located in the Three North area.By the end of 2014, the Three Northarea accounted for more than 80 percent of China’s total wind installationsbut its share of electricity consumption



is only 34 per cent. The uncoordinateddistribution of wind capacity requiresthe provision of electricity transmissionfrom wind-rich areas to load centres.However, China’s grid planning hasenjoyed neither comprehensivecoordination nor a long-term predictionof the rapid growth of wind energy.As a consequence, not only has theweakness of China’s transmission anddistribution infrastructure limited thetrade in wind power from large-scalewind projects to the developed cities atprovincial level, but it has also createdbarriers which affect wind powerconnections to the transmission grid atcountry level.

Secondly, current regulation modesfail to take wind power integration intoconsideration. Overall provincial yearlydispatch planning is mainly based onthe balance in each province, whileinter-provincial trade is mainly forgrid-security reasons. As thegeneration fleet is still dominated bycoal-fired power plants, the annualminimum generation quotas areallocated among coal-fired unitsthrough a highly political process, withless consideration being given to windpower generation at the beginning ofthe year, when the long-term predictionfor wind energy is inaccurate.Agreements relating to wind power areoverlooked in favour of the output ofcoal-fired power, as the annualminimum quotas should beguaranteed. In addition, a dailydispatch plan is usually set by thedispatching centre 24 hours ahead,when the accuracy of wind prediction islimited. That puts pressure on both thedispatching centre and coal-fired unitsif there is a need to decrease planned

coal power output for the purpose ofwind integration.

Thirdly, current related policies,especially feed-in tariffs, areunfavourable for the promotion oflarge-scale wind integration, as theylack incentives for other stakeholders.Currently, the feed-in tariffs for windpower are divided into four fixedcorresponding benchmarks accordingto wind resources only:

0.47 RMB/kWh,

0.50 RMB/kWh,

0.54 RMB/kWh and

0.60 RMB/kWh.

The implementation of feed-in tariffsfor wind power has played a significantrole in the development of wind energy,with the guaranteeing of wind farminvestors’ income at the early stagewhen wind generation only accountedfor a small share of electricityconsumption.

However, with the rapid growth of windinstallations, the inflexible pricingmechanism has had a negativeimpact, while benefiting otherstakeholders in the system, such aselectricity grid operators and coal-firedpower. The completely fixed tariff hasnot enabled any further large-scaleexpansion of wind energy utilizationsince the implementation of third-generation technology in the windsector; wind power utilization lagsbehind that of coal power and hydropower.

From the perspective of wind powertransmission among provinces, asthe wind feed-in tariffs are higher thanthe tariff for coal power, there is noeconomic incentive for neighbouringprovinces, especially the load centreprovinces, to promote the large-scaleimport of wind energy. Moreover,the completely fixed tariffs cannotreflect the advantage of wind powerin terms of low marginal generation

cost. In wind-rich areas, the increasingpenetration of wind fails to reduce thespot price of electricity through pricecompetition based on marginal cost,so it is difficult to find the true value ofgrid expansion and interconnectionutilization for other areas.

From the coal-fired power perspective:as integration of wind increases,coal-fired power generation declines.With the absence of income transfermechanisms from wind power to coalpower, coal power units are likely tosubmit undervalued down-regulationability to the dispatching centre toavoid a sharp decline of their utilization.

Thus, for grid operators, the currentpolicy framework is unable toincentivize the flexible regulation ofcoal-fired plants. From the perspectiveof the remaining participants in theenergy system, the fixed pricingpolicies on both the supply side andthe demand side do not enableflexibility in the existing system.When wind energy is curtailed, flexibleload (such as electric boilers and heatpumps) can technically enable theutilization of surplus electricity, butthe retail price cannot be reduced,so this fails to release any flexibility inthe system. Without suitablecompensation to flexible generatorsand consumers, the necessaryflexibility for large-scale wind utilizationin the existing system is constrainedand future improvements to flexibility(such as storage technologies anddemand-side management) will notattract the economic benefits whichwould allow technology and businessinnovations.

As a conclusion, the current policyframework, especially the fixed feed-intariff, is the main obstacle constrainingthe effective large-scale utilization ofwind power; consequently the system’sability to manage the uncertainty andvariability of wind energy is limited.

‘AGREEMENTS RELATING TO WIND

POWER ARE OVERLOOKED IN FAVOUR

OF THE OUTPUT OF COAL-FIRED

POWER…’



The failure of current government efforts toaddress wind curtailment

In order to improve the utilization ofwind energy, the Chinese governmenthas taken a series of actions.From 2013, the National EnergyAdministration (NEA) has requiredprovincial governments and regionalgrid companies to make windcurtailment mitigation a priority inwind development. And regional gridcompanies must disclose the latestinformation relating to wind utilization.In addition, the importance ofdevelopment in southern and easternparts of China was mentioned in 2014;this showed that future developmentshould consider not only the windresource but also the local electricityconsumption and grid structure.

‘FROM 2013, THE NEA HAS … MADE

WIND CURTAILMENT MITIGATION

A PRIORITY IN WIND

DEVELOPMENT.’

For grid construction, in 2015, theProgram for Strengthening thePrevention and Control of AtmosphericPollution (issued by the NEA)approved the construction of 12ultra-high voltage (UHV) channels tofurther expand the scale of thetransmission grid from north to southand from west to east. The UHVchannels were expected to relieve thebottleneck of wind exports from windfarms to load centres. For localizedwind power utilization, demonstrationprojects of wind heating are promotedto deal with wind curtailment innorth-east China.

In early 2013, the NEA proposed aplan supporting the replacement ofexisting coal boilers with electricboilers, by setting a specific electricityprice for each demonstration project,in order to use surplus wind energyfor space heating during winter. Forwind power storage, the NEA has

made an announcement suggestingthe acceleration of pumped hydrostation construction, for the purposeof providing potential regulationcapability for wind integration.

Government measures have aimedat decreasing wind curtailment,and it seemed as if their efforts hadseen some success in improvingwind utilization – from 2012 to 2014,curtailments in wind power were:20.8TWh, 16.2 TWh, and 14.1 TWh.However, these improvements incurtailment mitigation have beenlimited, and in the first six months of2015 the wind curtailment was 17.5TWh, which exceeded the annualcurtailment in 2013. Particularly in thewind-rich provinces (such as Gansu)the curtailment issue has deterioratedrapidly, with more than 60 per cent ofwind output in some wind farms wasrejected by the grid in 2015.

Power system reform would bring newopportunities to solve the problem

Wind power, as part of the electricitysystem in China, is dominated bythe administrative planning policyframework, which controls the quantityand price of electricity tightly. In abroad sense, China has succeededin expanding the generation andtransmission infrastructure needed tosupport its economic growth. However,the absence of market mechanismshas caused a failure to drive theelectricity system onto a more efficientand green track.

To encourage efficient marketcompetition, Beijing has ‘restarted theengine’ for power sector developmentby decoupling the transmission anddistribution tariffs from wholesaleprices, following a previous suspendedreform proposed in 2002.

In May 2015, the State Council releasedDocument 9, which issued the coreprinciple on further deepening the

reform of the power sector. The firsttask is to overhaul the role of gridoperators in the electricity marketand to set independent tariffs fortransmission and distribution, usingthe method of ‘cost plus reasonableprofits’. The core idea of the documentis to emphasize the role of marketmechanisms, injecting competitioninto both the generation and retailsegments of the industry through theoperation of a modern and practicalelectricity market. The other mainaspect of the reforms relates toassisting efforts to integrate renewableenergy from the perspectives of powersector decarbonization, air qualityimprovement, and climate changemitigation.

As mentioned above, the case of windcurtailment demonstrates thatpromoting the role of a market-basedpolicy framework in place of thecurrent administrative planningstructure is a vital approach forlarge-scale wind utilization. Based onthe market and an incentive-basedframework, spot market competitionwith flexible pricing would bringopportunities to solve the problemwhich can hardly be met in a heavilyadministrative planning system.

‘THE INTRODUCTION OF A SPOT

MARKET WOULD PROVIDE ESSENTIAL

PRICING SIGNALS FOR WIND ENERGY

TRADING.’

Firstly, the introduction of a spotmarket would provide essentialpricing signals for wind energy trading.In contrast to a fixed feed-in tariff,flexible wind pricing based on marginalcost can reflect the competitiveness ofwind generation that relates to lowvariable operation cost. If wind energyis able to decrease the spot electricityprice when wind resources are rich(such as in winter), neighbouring areaswould then be willing to importelectricity from wind-rich areas; this is



also preferable for wind power as it canthen sell it rather than curtail it. Fromthe long-term perspective, thedevelopment of wind power wouldhave an impact on the spot market,meaning that the concentration ofwind installations may increase theprice variation and decrease theaverage electricity price. Hence, theprospective wind investor wouldreconsider the location of wind farmsaccording not only to the windresource itself but also to thecorrelation between wind resourceand the local spot market prices. Inaddition, the spot market pricedifferences between regions wouldprovide the true value for gridexpansion, which helps to optimizethe grid structure and motivate theprovision of transmission channels.

Secondly, the formulation of marketand incentive-based mechanismshelps to remove reluctance on thepart of coal-fired power plant andgrid operators. Currently, greaterutilization levels have not received awelcome from coal-fired power plantand grid operators because thereis no reasonable compensation tocover their extra costs. Large-scalewind utilization would decrease thegeneration quota for coal power andattract extra costs when coal poweris needed to regulate for wind. With amarket-based framework, coal powerplants could get extra benefits throughpeak-regulation and security backup –mechanisms whereby the income forwind generators could more easily be

transferred to coal power plants. Thatwould also stimulate coal power plantsto improve the flexibility performanceof their units, in order to provide moreancillary services to the market.

Thirdly, market mechanismsmotivate the remaining stakeholdersto provide more flexibility for large-scale integration. According to theintermittency and volatility of windpower, the peak–valley price differencecould be boosted. When windgeneration is high, combined heat andpower (CHP) plants could decreasetheir power output and increase theirheat output, as the electricity pricewould be cheap. When wind generationis limited, CHP plants could increasetheir output of power at the expenseof heat, while any remaining heatingdemand is supplied by heat storage,as the spot price is expected toincrease. Thus, CHP plants – formerly abarrier to wind utilization in the heatingperiod – could now play a positiverole in ensuring wind utilization whenthere is enough wind and sufficientproduction capacity when there is nowind. Similarly, other stakeholders likedemand-side management, storagetechnologies, and electric vehiclescould also contribute to the balance ofthe electricity system.

In sum, the market-based frameworkproposed in Document 9 has thepotential to improve wind integrationsignificantly, on the basis of both abroad spot market and a flexible pricingsystem for all the generators.

In the existing game, grid operators,coal-fired power, and wind power wereall losing players without the incentiveto decrease wind curtailment. Withthe ongoing power reform, it could beseen that the construction of a modernelectricity market would promotecoordination between wind energy andthe rest of the energy system, which isthe long-term key to the realization oflarge-scale wind utilization.

The task of the reform is to find thenew equilibrium point for all the players.However, the capital cost of windpower is expected to be higher thanthat of coal power for the foreseeablefuture and protection for wind investorsis therefore required to avoid marketprice drops and wind-limited periods.To set suitable flexible prices for windpower, a combination of feed-in tariffand market price is suggested here, inwhich the actual return of wind energyis divided into two parts: the fixed windenergy subsidy and the market price.The feed-in tariff (the successful policyinstrument for wind promotion) shouldbe replaced by a feed-in premium(market price plus fixed wind tariff) orby another flexible pricing systemwhich combines the market pricesignal and subsidy.

‘THE MARKET-BASED FRAMEWORK

PROPOSED IN DOCUMENT 9 HAS

THE POTENTIAL TO IMPROVE WIND

INTEGRATION SIGNIFICANTLY. . .’



Christoph Burger is a faculty

member at the European School

of Management and Technology

(ESMT), Berlin

Scott Burger is a doctoral student at

the MIT Institute for Data, Systems

and Society

Rolando Fuentes is Research Fellow,

King Abdullah Petroleum Studies and

Research Center (KAPSARC), Riyadh

Iván Mario Giraldo is Economist,

Senior Associate (Latin America) at

Market Analysis Ltd

Richard Green is Professor of

Sustainable Energy Business at the

Business School, Imperial College,

London

David Harbord is Director at Market

Analysis Ltd

Tooraj Jamasb is Chair in Energy

Economics at Durham University

Business School, Durham

Malcolm Keay is a senior research

fellow at the Oxford Institute for

Energy Studies

Rabindra Nepal is a lecturer inEconomics at CDU Business School,Charles Darwin University, Darwin,Australia

David M. Newbery is Director of theEnergy Policy Research Group atCambridge University

Ignacio Pérez-Arriaga is VisitingProfessor at the Sloan School ofManagement (MIT) and Professorat Comillas University and theEuropean University Institute inFlorence

Rahmat Poudineh is lead researchfellow for the electricity programme,Oxford Institute for Energy Studies,Oxford

David Robinson is a senior researchfellow at the Oxford Institute forEnergy Studies and President, DavidRobinson & Associates

Fereidoon P. Sioshansi is President ofMenlo Energy Economics, California

Goran Strbac is Chair in ElectricalEnergy Systems at the Department ofElectrical and Electronic Engineering,Imperial College, London

Graham Weale is Chief Economist at

RWE AG, Essen

Jens Weinmann is Programme

Director at the European School

of Management and Technology

(ESMT), Berlin

Xin Li is a research fellow at the

Oxford Institute for Energy Studies

Xiong Weiming is a doctoral student at

the Institute of Energy, Environment

and Economy, Tsinghua University

Zhang Xiliang is Professor of Energy

Economics and Policy at the

Institute of Energy, Environment and

Economy, Tsinghua University

The views expressed in this issue are

solely those of the authors and do

not necessarily represent the views

of OIES, its members, or any other

organization or company.

CONTRIBUTORS TO THIS ISSUE



EDITOR: Bassam FattouhCO-EDITORS: Rahmat Poudineh, David Robinson, and Malcolm KeayAnnual Subscription (four issues) £90.00

Oxford Energy Forum. ISSN 0959-7727. Published by Oxford Institute for Energy Studies, Registered Charity 286084.

© Oxford Institute for Energy Studies, 2016.

57 Woodstock Road Oxford OX2 6FA

Direct Line: +44 (0)1865 889136Reception: +44 (0)1865 311377Fax: +44 (0)1865 310527

www.oxfordenergy.org

Recently published book

Europe’s Long Energy Journey –Towards an Energy Unionby David Buchan and Malcolm Keay

February 2016

Recent Research Papers

Under the Mountains – Kurdish Oiland Regional Politicsby Robin Mills

January 2016

Electricity Markets are Broken – Canthey be Fixed?by Malcolm Keay

January 2016

Fiscal Stabilization in Oil andGas Contracts – Evidence andImplicationsby Mario Mansour and Carole Nakhle

January 2016

The Evolution of European TradedGas HubsPatrick Heather

December 2015

Incidence and Impact – ADisaggregated Poverty Analysis ofFossil Fuel Subsidy Reformby Jun Rentschler

December 2015

Reforming Electricity Reforms? EmpiricalEvidence from Asian Economiesby Anupama Sen, Rabrindra Nepal, and

Tooraj Jamasb

February 2016

Natural Gas in China – A RegionalAnalysisby Xin Li

November 2015

A Holistic Framework for the Study ofInterdependence Between Electricityand Gas Sectorsby Donna Peng and Rahmat Poudineh

November 2015

US Shale Oil Dynamics in a Low PriceEnvironmentby Trisha Curtis

November 2015

Oil in Uganda – Hard Bargaining andComplex Politics in East Africaby Luke Patey

October 2015

The Dynamics of the RevenueMaximization – Market Share Trade-off: Saudi Arabia’s Oil Policy in the2014–2015 Price Fallby Bassam Fattouh, Rahmat Poudineh,

and Anupama Sen

October 2015

NEW OIES PUBLICATIONS

Recent Energy Comments

Gazprom – Is 2016 the Year fora Change of Pricing Strategy inEurope?by James Henderson

January 2016

Has the North Sea Entered a Late-lifeCrisis?by Virendra Chauhan, Maarten van

Mourik, and Amrita Sen

December 2015

The New Economics of Oilby Spencer Dale

October 2015



218964 chromalin 20160216155100the business model of utilities, new business models can shape future...

Documents