skip to content

Timetable (MESW02)

Electricity systems of the future: incentives, regulation and analysis for efficient investment

Monday 18th March 2019 to Friday 22nd March 2019

Monday 18th March 2019
09:15 to 09:35 Registration
09:35 to 09:45 Welcome from David Abrahams (Isaac Newton Institute)
09:45 to 10:45 David Newbery
Market and regulatory design for renewables dominated systems
low-carbon electricity system has a dominance of plant with high fixed costs
and low variable costs, backed up with flexible controllable generation. In a
liberalised market all required new investment will need to cover its full
cost, necessitating payments for capacity, energy and quality of service to
deliver reliability (long-term adequacy) and security of supply (short term
resilience to shocks). Their value varies over time and space. Current
liberalised markets lack futures markets and proper spatial signals to guide
investment. Networks have always had high fixed, very low variable costs and
long-run marginal costs well below average costs, creating challenges for
setting network tariffs. Consumers need to pay for these generation and
network services in ways that are efficient, equitable and acceptable, in a
system that offers a greater range of scale and location (voltage level) of
options for delivering the three services of capacity, energy and quality of
servic e, requiring more careful tariff design than hitherto. ICT can help
deliver but cannot avoid the tensions between efficiency, equity and
acceptability. The talk will identify the challenges, the theoretical
solutions drawing on the mature subject of public economics, and possible
10:45 to 11:15 Morning Coffee
11:15 to 12:00 Natalia Fabra
Competition among Renewables
We model strategic behavior of renewable suppliers when competing in electricity auctions. We introduce renewables' intermittency by assuming that firms' available capacities are random and private information. In equilibrium, bid functions are a smooth decreasing function of firms' realized capacities. Thus, at times when there is more renewables' availability, supply functions shift outwards and downwards, leading to reductions in the market price. An increase in correlation between firms' available capacities strengthens competition non-monotonically. Keywords: electricity, competition, auctions.Natalia Fabra and Gerard Llobet (CEMFi)
12:00 to 12:45 Bert Willems
Investments in flexible generation capacity in an energy-only market
A de-carbonized energy system will require sufficient investments in flexible generation capacity. The goal of this paper is to study how market design affects investments in flexible generation capacity. If markets are perfectly competitive, production costs are convex, and demand is perfectly predictable, then independent energy-only market (one for each hour of operation) will lead to efficient market outcomes both in the short-run (market operation) and in the long-run (investment levels and levels of flexibility).We will relax the assumption of convex production costs and predictable demand and analyze first whether small competitive investors have the right incentives to invest in flexible generation. In a second step we intend to derive equilibrium investment levels.
12:45 to 13:45 Lunch at Churchill College
13:45 to 14:30 Richard Green
Optimal storage, investment and management under uncertainty - It is costly to avoid outages!
We show how electricity storage is operated optimally when the load net of renewable output is uncertain. We estimate a diurnal Markov-process representation of this residual load in Germany in 2011 to 2015 on an hourly basis and design a simple dynamic stochastic electricity system model with non-intermittent generation technologies and storage. We derive the optimal storage, generator output and capacity levels. If storage capacity replaces some generation capacity, the optimal storage strategy must balance arbitrage (between periods of high and low marginal cost) against precautionary storage to ensure energy is available throughout a long peak in net demand. We then solve the model numerically under realistic conditions and compare the results to perfect foresight findings. We show that a perfect foresight model would over-estimate the cost-saving potential of energy storage by 18%, as it takes up arbitrage opportunities that ignore the need for precautionary storage.Joint work with Joachim Geske.
14:30 to 15:15 Pär Holmberg
Capacity mechanisms and the technology mix in competitive electricity markets
The paper is co-authored with Robert Ritz at Judge Business School and EPRG, University of Cambridge.

Capacity mechanisms are playing a growing role as part of electricity market design in Europe, North America and other jurisdictions. Yet their role remains hotly debated with some electricity systems retaining an "energy-only" market design without apparent need for capacity payments. In this paper, we introduce a new model of a capacity mechanism in a market with a continuum of generation technologies. We consider two policy instruments: a wholesale price cap and a capacity payment (or procured capacity volume). We show that some combinations of policy instruments will result in socially optimal market investments. Changing capacity payments and the price cap will only influence investments in peak generation plants. Investments improve system reliability, which is a public good. We find that capacity payments can be used to internalize this externality. We also find that capacity payments can be used to mitigate market power for a given social welfare level.
15:15 to 15:45 Afternoon Tea
15:45 to 16:30 Eddie Anderson
The role of submodularity in capacity auctions
We consider a capacity auction in which multiple supplying firms offer bids, and an agent, the buyer, selects which bids to accept. Payments may depend on the set of bids accepted. We consider the role played by a submodularity property for the social welfare function in terms of the set of participating firms. We show how submodularity leads to good properties for the equilibrium. We illustrate this through discussion of a case where the buyer faces uncertain demand and there are separate costs incurred by the suppliers for making capacity available (reservation costs) and for delivering against the required demand (execution costs). We demonstrate that when marginal costs are constant the submodularity property holds, and in equilibrium each supplier makes a profit equal to their marginal contribution and the overall expected welfare is maximized. Eddie Anderson (Lusheng Shao and Bo Cheng)
16:30 to 17:15 Hannes Weigt
Long Term Electricity Market Design: Pricing Quality?
Future electricity market design needs to address a set of challenges to provide for efficient investment, supply and consumption decisions. The objective of this paper is to sketch those problems and challenges embedded along the electricity value chain and identify how they are altered when shifting from a pre-dominantly centralized nuclear and fossil system towards a system with a mix of central and decentralized elements as well as a high share of intermittent renewable generation. Of particular concern is the mismatch of incentive structures (i.e. between highly dynamic wholesale markets setting incentives for large scale investors and average tariff structures on the end user side setting incentives for prosumers) as well as the security of supply dimension (i.e. shifting from a public good character within regulated systems towards a potential private good in a system characterized by smart control structures). Based on those structures we aim to derive potential market and price setting designs suited for a future electricity market. Frank C. Krysiak and Hannes Weigt
17:15 to 18:15 Welcome Wine Reception at INI
Tuesday 19th March 2019
09:00 to 10:00 Richard O'Neill
Incentives, Regulation and Analysis for Investments in ISO Markets
The presentation will examine the history of US Independent System Operator (ISO) markets, mistakes that were made and corrected, how they are regulated, and how theoretical sustainable economic efficiency can be achieved. In addition, we examine how efficient transmission planning in theory is achieved and the problems associated with implementing it. Lastly, we will address the critical role optimization software and approximations must play to achieve these objectives.
10:00 to 10:45 Sarah Ryan
Assessing Potential Benefits of Increased Gas-Electric Coordination by Stochastic Optimization
Natural gas is the single largest fuel source for electricity generation in the US. Its share of electric energy generated is expected to increase through 2050 while the shares attributed to coal and renewable sources approximately swap their values. However, electric power generation accounts for only about a third of gas consumption and this fraction is expected to remain nearly constant. Despite some recent regulatory changes, the markets for gas and electricity are independent and not well coordinated. This lack of coordination combined with uncertainty associated with variable renewable generation creates risks for both the generators that procure gas by interruptible contracts and the system operators charged with maintaining both reliability and low wholesale electricity prices. We formulate stochastic programming models for daily unit commitment and dispatch with uncertain wind generation to represent operation both within the current uncoordinated markets and under a hypothetical integration of the gas and electricity systems. Comparison of cost and reliability metrics across models allows an estimation of the potential benefits of increased coordination between the two systems.Sarah RyanandDan Hu
10:45 to 11:15 Morning Coffee
11:15 to 12:00 Ramteen Sioshansi
Merchant Storage Investment in a Restructured Electricity Industry
Restructuring and liberalisation of the electricity industry creates opportunities for investment in energy storage, which could be undertaken by a profit-maximising merchant storage operator. Because such a firm is concerned solely with maximising its own profit, the resulting storage-investment decision may be socially suboptimal (or detrimental). This paper develops a bi-level model of an imperfectly competitive electricity market. The modelling framework assumes electricity-generation and storage-operations decisions at the lower level and storage investment at the upper level. Our analytical results demonstrate that a relatively high (low) amount of market power in the generation sector leads to low (high) storage-capacity investment by the profit-maximising storage operator relative to a welfare maximiser. This can result in net social welfare losses with a profit-maximising storage operator compared to a no-storage case. Moreover, there are guaranteed to be net social welfare losses with a profit-maximising storage operator if the generation sector is sufficiently competitive. Using a charge on generation ramping between off- and on-peak periods, we induce the profit-maximising storage operator to invest in the same level of storage capacity as the welfaremaximising firm. Such a ramping charge can increase social welfare above the levels that are attained with a welfare-maximising storage operator.Afzal S. Siddiqui, Ramteen Sioshansid and Antonio J. Conejo
12:00 to 12:45 Ben Hobbs
Sources and Implications of Inaccuracies in Capacity Credit Calculations: A Static Analysis of Electric Generation Capacity Markets
A static (single shot) model of capacity investments for thermal and renewable resources under energy and capacity markets as well as energy price caps is considered. Under assumptions of risk neutrality and continuous capacity investment variables, it is proven that there exists a capacity (in $/MW capacity credit/yr) and a set of capacity credits by generation type (MW credit/MW nameplate capacity) that supports the most efficient mix of capacity types. However, capacity credits are often set for political reasons, fail to reflect dynamics of levels of penetration upon the marginal contribution of capacity to reliability, or are estimated considering inadequate sample sizes of variable energy output. Through market simulations for an ERCOT-like system, the genmix and efficiency impacts of credits deviating from first-best levels are calculated for various scenarios of erroneous credits.
(Co-author: Cynthia Bothwell, US Dept. Energy, Office of Energy Efficiency & Renewable Energy)
12:45 to 13:45 Lunch at Churchill College
13:45 to 14:30 Claudia Sagastizabal
A Two-Stage Model for Planning Energy Investment under Uncertainty
We consider risk-averse stochastic programming models for the Generation and Expansion Planning(GEP) problem with investment decisions in the first stage and generation variables of recourse, decided in a second stage. The resulting problem is coupled both along scenarios and along power plants. To achieve decomposition, we combine the Progressive Hedging approach in [1] with a suitable duplication of variables [2]. The resulting nonsmooth dual function can then be solved with an inexact dual proximal bundle method, as in [3]. The procedure defines a primal-dual sequence that, under reasonable assumptions, is shown to converge to a primal-dual solution of the original problem.Claudia Sagastizabal andF. Atenas
14:30 to 15:15 Anthony Papavasiliou
Market Design Considerations for Scarcity Pricing: A Stochastic Equilibrium Framework
We develop a stochastic equilibrium framework for analyzing variations of two-settlement system in which energy and reserve capacity is traded in a day-ahead market followed by a real-time market. The framework is aimed at analyzing the impact of various short-term market design decisions on the remuneration of reserve capacity under operating reserve demand curves. The proposed framework accounts for risk aversion, real-time uncertainty, and a relaxed representation of unit commitment decisions. The framework can be used for analyzing the implication of various market design choices on the back-propagation of real-time prices. These choices include (i) the presence or absence of a real-time reserve capacity market, (ii) the simultaneous or sequential clearing of reserves and energy in day-ahead markets, and (iii) the presence or absence of virtual bidding. We propose a decomposition heuristic for solving the resulting non-convex equilibrium problem. We apply our framework for the analysis of market design choices on the remuneration of reserves in the Belgian electricity market.
15:15 to 15:45 Afternoon Tea
15:45 to 16:30 Jong Shi Pang
Two-stage Stochastic Programming with Linearly Bi-parameterized Quadratic Recourse
This paper studies the class of two-stage stochastic programs (SP) with a linearly bi-parameterized recourse function defined by a convex quadratic program. A distinguishing feature of this new class of stochastic programs is that the objective function in the second stage is linearly parameterized by the first-stage decision variable, in addition to the standard linear parameterization in the constraints. Inspired by a recent result that establishes the difference-of-convexity (dc) property of such a recourse function, we analyze the almost-sure subsequential convergence of a successive sample average approximation (SAA) approach combined with the difference-of-convex algorithm (DCA) for computing a directional derivative based stationary solution of the overall non- convex stochastic program. Under a basic setup, the analysis is divided into two main cases: one, the problem admits an explicit, computationally viable dc decomposition with a differentiable con- cave component, based on which the discretized convex subproblems to be solved iteratively can be readily defined; and two, an implicit bivariate convex-concave property can be identified via a certain smoothing of the recourse function. The first case includes a strictly convex second-stage objective and a few special instances where the second-stage recourse is convex but not strictly convex. A general convex second-stage recourse function belongs to the second main case; this case requires the introduction of the notion of a generalized critical point to which the almost-sure subsequential convergence of the combined SAA and DCA is established. Overall, this research provides the first step in the investigation of this class of two-stage SPs that seemingly has not been, until now, the object of a focused study in the vast literature of computational two-stage stochastic programming.
16:30 to 17:15 Luiz Augusto Barroso
Harmonizing energy planning and market mechanisms to ensure supply adequacy in electricity markets
The objective of this talk is twofold: we
first describe multi-stage stochastic optimization models with a very
detailed representation of the system components that have been used to carry
out realistic generation-transmission planning studies in large systems with
strong renewable penetration. The optimization framework of these models
allows not only to extract meaningful outputs - such as probabilistic-reserve
needs, cost of flexibility - as well as they can be used to factor planning
outcomes in auctions for long-term reliability products to ensure supply
adequacy. We will show that attributes of generation plants can be factored
in the procurement process of these products via primal (price-based) or dual
(quantity-based) outcomes of the optimization-based framework for the
planning studies.
Wednesday 20th March 2019
09:00 to 10:00 Sonja Wogrin
Hierarchical optimisation and equilibrium problems in electricity systems: challenges and status quo
talk starts off with a brief introduction to hierarchical optimisation and
equilibrium problems, and continues with a broad overview of different
applications of these types of problems within the electricity system (such
as transmission and generation expansion planning, strategic bidding in
markets and many more). We then establish the status quo of the existing
models and analyse: what level of technical detail existing models cover;
what are corresponding CPU times, model size and complexity; what are the
current methods available to solve the arising problems, and, raise the
discussion of what gets lost in translation. With the former analysis in
mind, we outline the existing lack in the literature, or better yet the
existing need to represent complex technical realities in our models. Complex
realities such as: the AC optimal power flow, or unit commitment constraints.
We then try to establish the link between the technical nature, and the
corresponding arising mathematical difficulties in the hierarchical models,
discussing issues such as non-convexities and non-existence of solutions. The
talk concludes by pointing out some of the most recent methodological
advances in these directions, and what still remains to be done.
10:00 to 10:45 Chris Dent
Capacity markets, risk modelling and decision support
talk will introduce the mathematical and statistical questions associated
with running a capacity market. It will then summarise recent research on the
modelling of demand and wind generation for capacity adequacy risk assessment
in Great Britain, and on decision criteria for determining the volume to
procure. Finally, it will describe a new theory of how to run a capacity
auction including all resources including storage on a common basis. All
these topics will be illustrated using examples based on the GB system.
10:45 to 11:15 Morning Coffee
11:15 to 12:00 Stein-Erik Fleten
The Effect of Capacity Payments on Peaking Generator Availability in PJM
study the effects of capacity payments on the strategic decisions of plant
managers for peaking units in the PJM Interconnection. We achieve this
through a structural estimation of maintenance and switching costs between
the operational state, the standby state and retirement of generating units,
using annual data from 2001-2016. The empirical data shows less switching
between states after the introduction of capacity remunerations in 2007. We
find that the role of peaking units has changed, with the units being
dispatched more often. In the counterfactual analysis, we find a clear
connection between the level of capacity payments and switching. We conclude
that the current level of capacity payments in PJM incentivizes peaking units
to stay in the operational state.Stein-Erik Fleten (with Benjamin Fram, Magne Ledsaak, Sigurd Mehl, Ola Røssum, Carl Ullrich)
12:00 to 12:45 Asgeir Tomasgard
Analysing effects of short- and long-term uncertainty on capacity expansion in European electricity markets
EMPIRE model is a European multiscale power market model with investments
towards 2050 as well as representative hours. It is well suited to capture
operational uncertainty in generation from intermittent energy sources like
wind and sun. In the short-run
horizon, typical uncertainty is
in load, intermittent generation and inflows to hydro reservoirs. Short-run
sources of flexibility are regulated hydropower, storages, fossil generators
and demand response. In the long run uncertain factors include learning curves, policy uncertainty,
long-term commodity process and demand trends. In this paper we add long-term
uncertainty to the formulation. The resulting models are large scale
stochastic multi-stage recourse models with hundred of millions of variables.
We present both solution methods and analysis of the most important factors.
In the short-run horizon, typical
uncertainty is in load, intermittent
generation and inflows to hydro reservoirs. Short-run sources of flexibility
are regulated hydropower, storages, fossil generators and demand response. In
the long run uncertain factors include
learning curves, policy uncertainty, long-term commodity process and
demand trends. We look at the European level and analyze questions like to what extent the demand response potential
can facilitate an optimal transition to an European low emission power system
and how does long-term uncertainty affect the investments in renewables.
12:45 to 13:45 Lunch at Churchill College
13:45 to 18:00 Free afternoon
19:30 to 22:00 Formal Dinner at Westminster College

Westminster College


Open Camembert, fig and onion tart with frisée salad and a Balsamic glaze.  
Quinoa, rocket, pomegranate and heritage tomato salad with a lemon and Balsamic dressing (Vegan)  
Main course
Guinea fowl breast with black pudding stuffing, fondant potato, roasted butternut squash, apple purée, kale, baby turnips and a Madeira jus.  
Roasted vegetable stack, sautéed Jerusalem artichokes, butternut squash purée, split lentils, salsa verde and a Brussel sprout and sun-dried tomato salad (Vegan)  
Eton mess cheesecake, with strawberry coulis and a fruit of the forest compote.

Smart casual
Thursday 21st March 2019
09:00 to 10:00 Yves Smeers
Risk premium and stochastic equilibrium in generation capacity expansion models
in new capacities is most often based on a Weighted Average Cost of Capital
where the cost of equities is derived from the Capital Asset Pricing Model
(the cost of debt following a different logic). While a single corporate
discount rate was most often used in the past, differentiated discount rates
reflecting elements such as technology and country risks are now more used.
These introduce undesirable arbitrage phenomena in standard capacity
expansion model interpreted as market simulation models. We discuss three
types of differentiated discounting namely the standard (often based on CAPM)
exogenous discount rate, the more general stochastic discount rate and the
endogenous discount rate derived from risk functions possibly with hedging
instruments. The three approaches are cast in a unifying risk premium
equilibrium formulationwhere arbitrage phenomena are eliminated. The
models are of the complementarity form and can be handled through splitting
algorithms. We report numerical results for medium size problems adequate for
industrial use. Convergence is based on an(often implicit)
assumption of monotonicity that is not necessarily satisfied for endogenous
discount rates. We briefly discuss the different equilibrium that can arise
when this assumption is not satisfied.
10:00 to 10:45 Danny Ralph
Risky Capacity Equilibrium Models for Risk Averse Investment Equilibria with Incomplete Markets
"Risky Capacity Equilibrium Problems” incorporate (i) risk averse investment in power plants, (ii) financial trading to hedge those investments, and (iii) strategic production in a stochastic spot market. These models concatenate short-term electricity market (perfect competition or Cournot) with long-term investments (risk neutral or risk averse behaviour in different risk trading settings). We focus on incomplete financial markets, when not all risks can be traded, using results on “Risky Design Equilibrium Problems” and standard Nash game techniques to show existence of equilibria. Numerical results show the impact of incompleteness on equilibrium capacity and spot prices.
10:45 to 11:15 Morning Coffee
11:15 to 12:00 Michael Ferris
Modelling 100 percent renewable electricity
Michael Ferris and Andy Philpott
12:00 to 12:45 Gauthier de Maere d'Aertyrcke
Valuation of floating price contract formulae for financial renewable PPAs
We propose a valuation methodology based on risk measure to compute the risk premium to cover new structure of financial renewable PPA, where as opposed to traditional unit contingent PPA, a part of merchant risk remains in the hand of the asset developer. Those new contract structures are currently emerging on the market: Offtaker are increasingly showing interest to share the market risk, as in systems with high RES share, the PPA might become a very poor hedge of their electricity bill (due to RES cannibalization effect). We specifically focus on contract formulae based on floating price, where the financial transfer to the Offtaker is not based on the price captured by the renewable asset but rather on a floating price, defined as the average of spot prices over a predefined time duration (e.g. yearly, monthly or daily basis). We then assess several risk mitigation strategies to lower the risk premium, namely physical (well balanced portfolio of wind, PV and battery) and/or financial (power derivatives trading).
12:45 to 13:45 Lunch at Churchill College
13:45 to 14:30 Jalal Kazempour
Distributionally robust chance-constrained generation expansion planning
This talk addresses a centralized generation expansion planning problem, accounting for both long- and short-term uncertainties. The long-term uncertainty (demand growth) is modeled via a set of scenarios, while the short-term uncertainty (wind power) is considered using moment-based ambiguity sets. In the expansion stage, the optimal units to be built are selected among discrete options. In the operational stage, a detailed representation of unit commitment constraints is considered. To make this problem tractable, we solve it in linear decision rules, and use a tight relaxation approach to covexify the unit commitment constraints. The resulting model is a distributionally robust chance-constrained optimization, which eventually recasts as a mixed-integer second-order cone program. We consider the IEEE 118-bus test system as a case study, and explore the performance of the proposed model using an out-of-sample analysis.
14:30 to 15:15 Jacob Mays
Asymmetric Risk and Fuel Neutrality in Capacity Markets
This paper calls into question the fuel neutrality of capacity mechanisms implemented in liberalized electricity markets. The argument relies on two assumptions likely satisfied in practice, first that investors are risk averse and second that markets in risk are incomplete. For the analysis, we develop a heuristic algorithm to solve large-scale stochastic equilibrium models describing a competitive market with incomplete risk trading. Introduction of a capacity mechanism has an asymmetric effect on the risk profile of different generation technologies, tilting the resource mix toward those with lower fixed costs and higher operating costs. One implication of this result is that current market structures may be ill-suited to financing low-carbon resources, the most scalable of which have high fixed costs and near-zero operating costs. Development of new risk trading mechanisms to replace or complement current capacity obligations could lead to more efficient outcomes. Jacob Mays, David Morton and Richard O'Neill
15:15 to 15:45 Afternoon Tea
15:45 to 16:30 Golbon Zakeri
Stochastic auctions to accommodate the future flexi-grid
We will
discuss the design and performance of somestochastic auctions designed
to accommodate flexibility and hence penetration of intermittent renewable
energy generation. We will examine these auctions in the context of
competitive markets as well as oligopolies. We will also shed some light on
using such mechanisms in markets where the participants are risk averse. The
talk is based on the following literature.
16:30 to 17:15 Michael Hintermüller (Weierstraß-Institut für Angewandte Analysis und Stochastik); (Humboldt-Universität zu Berlin)
Generalized Nash Equilibrium Problems with Application to Spot Markets with Gas Transport
A class of noncooperative Nash equilibrium problems is presented, in which the feasible set of each player is perturbed by the decisions of their competitors via a convex constraint. In addition, for every vector of decisions, a common “state” variable is given by the solution of an affine linear equation. The resulting problem is therefore a generalized Nash equilibrium problem (GNEP). The existence of an equilibrium for this problem is demonstrated, and first-order optimality conditions are derived under a constraint qualification. An approximation scheme is proposed, which involves the solution of a parameter-dependent sequence of standard Nash equilibrium problems. An associated path-following strategy based on the Nikaido–Isoda function is then proposed. Function space- based numerics for parabolic GNEPs and a spot-market model are developed.
Friday 22nd March 2019
09:00 to 10:00 Andrew Wright
Societal objectives and their impact on the electricity system of the future
Electricity is an essential to people’s well-being and the lifeblood of a modern economy. The societal objectives around electricity reflect the wider values of society, and these values precondition the way the electricity system is structured and governed, both today and in the future. The electricity system is changing rapidly, but societal values are usually more enduring, and will shape the ways in which the system develops. The challenge of the energy system of the future is use technology to meet the goal of decarbonisation, without undermining the reasonable expectations of society for an essential public service. This talk will look at these societal objectives, how they shape the electricity system of today and the risks and opportunities in the transition to a low carbon future.
10:00 to 10:45 Goran Strbac
Energy system decarbonisation: informing reform of energy market and regulatory framework
There is a growing need and opportunity for new emerging technologies and systems to provide flexibility in supporting a cost-effective transition to a lower carbon energy system. Key results of the comprehensive analysis carried out will be presented highlighting the benefits of emerging technologies and system in providing grid support, various balancing services, system security services and option value services to deal with uncertainty. This will underline the barriers related to the present market and regulatory framework that may prevent the realization of the quantified system value of flexibility and thus weaken the business case for investment in corresponding technologies and systems. Developing fully cost-reflective markets and effective regulatory frameworks will therefore be necessary to ensure that the commercial incentives for investing and operating flexible technologies are aligned with the societal benefits. The importance of the whole-system approach to future market design, strategic versus incremental investment, consumer participation and decentralisation will be discussed and a roadmap for market evolution will be presented, highlighting the importance of aligning the market design with the decarbonisation objectives.
10:45 to 11:15 Morning Coffee
11:15 to 12:00 Afzal Siddiqui
Strategic Storage Use in a Hydro-Thermal Power System with Carbon Constraints
The Northeast Power Coordinating Council (NPCC) comprises American states and Canadian provinces marked by a significant penetration of variable renewable energy sources (VRES) and hydropower production. Major demand centres in New England, New York, Ontario, and Québec that are subject to stringent to stringent caps on CO2 emissions are included in the NPCC. For example, the Regional Greenhouse Gas Initiative (RGGI) mandates a 30% reduction in CO2 emissions from power plants by 2030 relative to 2020 levels, which affects generation in New England and New York. Likewise, Québec participates in the Western Climate Initiative (WCI), which aims to reduce CO2 emissions by approximately 40% by 2030 relative to 1990 levels and included Ontario until recently. Both RGGI and WCI create cap-and-trade (C&T) systems for CO2 emissions in which the shadow price on the binding CO2 emission constraint is the permit price that generators incur as an additional cost for their CO2 emissions. While support schemes such as feed-in tariffs and the C&T system have induced an increase in VRES generation, they have also enhanced the role of energy storage, viz., by hydro reservoirs especially in Québec. In a perfectly competitive power system, storage capacity would be deployed in a socially optimal way to smooth out the fluctuations in uncontrollable VRES output. However, given the persistence of market power in the electricity industry, hydro reservoirs may be used in a strategic manner to the benefit of their proprietors. Consequently, incentives for VRES and social welfare may be detrimentally affected by such exertion of market power. In order to investigate the extent of these distortions in the NPCC and to propose policies for their mitigation, we develop a bottom-up equilibrium model to quantify the welfare losses from the strategic use of hydropower reservoirs and to assess counterfactual CO2 emission caps.
Co-authors: Sébastien Debia and Pierre-Olivier Pineau
12:00 to 12:45 Duncan Burt
Risk and Optimization in Future Network Planning
TThe future of power network use is highly uncertainty, with the arrival of renewables and new patterns of power use, creating large uncertainties, risks and opportunities for the optimisation of future investment and development.  National Grid runs an annual Network Options Assessment which applies a risk-based economics approach to optimise which investments are progressed on an annual basis. his involves modelling a complex power grid and the timing and choice of many potential investment options, totalling multiple billions of pounds over the next 30 years, to deliver an efficient and secure network. This talk sets out the challenges and key factors involved in our current optimisation process and seeks views on alternative approaches.
12:45 to 13:45 Lunch at Churchill College
13:45 to 14:30 Amy Wilson
Accounting for uncertainty when using computer models as decision-support tools in energy system planning
models are widely used as decision-support tools for planning energy systems
in both industry and government. These computer models are often
computationally intensive and have high-dimensional input spaces, making it
difficult to quantify the impact that different sources of uncertainty have
on model output. Without a complete picture of the effect of these
uncertainties it is difficult to take planning decisions that are robust in
the real-world. This presentation will discuss methodology for accounting for
uncertainties in computationally intensive energy planning models. Both input
uncertainty and uncertainty in the structure of the model itself will be
considered. An emulator, or statistical model of the underlying computer
model, will be used to quantify uncertainty in areas of the input space where
it has not been possible to make model runs. This emulator will be combined
with a description of the uncertainty over the input space and a description
of the structural error to quantify uncertainty in model outputs. Several
real-world examples in energy planning will be discussed, including the
modelling of wholesale electricity prices and making decisions about
renewable support schemes.
14:30 to 15:15 Jean-Paul Watson
On the Rigorous Evaluation of Stochastic Approaches to Power Systems Operations
there is significant recent research on stochastic optimization approaches to
power systems operations, e.g., unit commitment and economic dispatch, there
are still major impediments to their adoption in practice. In our experience,
developed over years of attempting to deploy such approaches, one key issue
is accurate evaluation of any proposed approach, relative to existing
deterministic operational methodologies. In this talk, we discuss the
challenges in such evaluation, and report on a novel methodology addressing
what we feel to be deficiencies with current approaches. In the talk, we
focus on issues relating to data availability and segmentation, probabilistic
scenario generation, and the impact of scenarios on operational performance.
15:15 to 15:45 Afternoon Tea
15:45 to 16:30 Andy Philpott
Long-term generation capacity expansion models in JuDGE
consider a multistage stochastic programming model of capacity expansion for
renewable energy in New Zealand. The model is solved using the JuDGE package
in Julia that exploits a Dantzig-Wolfe decomposition of the problem.Andy Philpott and Anthony Downward
University of Cambridge Research Councils UK
    Clay Mathematics Institute London Mathematical Society NM Rothschild and Sons