introduction to electricity markets

~ markets as optimization systems ~

  • questions:
    • where does power come from?
    • who decided, and when, which generators in alberta would be running at this minute?
    • what did the last megawatt hour cost ot produce? what was it sold for? what are the reasons for differentiation at two different points in time?
  • why study electircity markets?
    • ‘essential intermediate good’
      • quanitites balanced every second
      • prices computed on minute scale
      • optimization across a variety of periods
      • extremely complex economic ecosystem
    • unusual economics
      • limited storage at scale
      • demand barely responds to price
      • power system reliability is a public god (non rivalrous and non-excludable)
      • flows on power networks governed by physics rather than contracts or policy
      • alberta; energy only market since 2001, subject to continuous evolution
      • REM: restructured energy market
        • chnages how alberta dispatches and prices power
        • modern electricity market: REM subset
    • 'a wholesale electricity market is a decentralized mechanism for solving a constrained cost-minimization problem to efficiently allocate resources'
      • ‘wholesale’-selling product in bulk to other businesses rather than to individual consumers
      • ---> having electricity sold wholesale and the structure through which its sold allows for decentralization of a solution to a cost-minimation problem under specific constraints so that consumers can have an intermediate essential good that should be efficiently allocated
      • market design feature will be one of three things:
        • an objective
        • a constraint
        • rule for turning the solution into prices and allocating resources
  • the physical system
    • generation: production of electricity
      • hundreds of facilities in alberta
    • transmission: high voltage grid moving bulk power across the province
    • distribution: lower voltage wires reaching premises
      • regulated monopolies by service territory
    • retail: consumer interface with the market
  • uniqueness of electricity
    • limited storage at commericial scale: network downtime means balckout/brownouts
      • ‘brownout’?
      • how would a network go down?
    • equilibirum reached by moving supply
      • automated price responsiveness of demand and storage at scale makes this even more complex
    • delivery follows physics, not contracts or policy
      • location matters if the network is constrained
        • physical mechanism versus societal construct — how does this change things?
    • capital intensity and longevity:
      • generators expensive and lumpy investments that last for decades
        • today’s prices and expectations are revenue that underpin investment
  • timeline
    • twenty years ago:
      • system operator projects long term transmission plan over a twenty year horizon
    • five to ten years ago:
      • S.O applies to the regulator for a new transmission:
      • generators begin considering new investment, approval and build times of one to ten years:
      • investment case requires a forecast of energy market prices over time
    • two to five years ago:
      • construction begins on new generation and transmission:
      • generators declare outages for maintenance:
      • generators, large consumers, and retailers begin hedging in the forward market
        • ‘forward market’: over the coutner financial marketplace, two parties agree to buy or sell and asset at a set price for delivery at a specific time
        • ‘hedging’: taking financial action to protect against the risk of future price increases
    • one day ahead:
      • SO procures operating reserves for the following day: capacity held back from the energy market for contingencies and moment-to-moment balancing
      • generators submit offers 999.99/MWh
      • importers offer to bring power in with interconnections with BC, montana, SK: exporters bid to take the power out
      • consumers can submit bids to reduce consumption at stated price, although this is rare
      • all offers and bids are assembled into a merit order for each hour: the ordered supply curve this course studies
        • merit order: feasibility? how realistic it is? what good it does?
    • up to two hours ahead:
      • generators can change (restate their offer prices)
      • offers for the hour beginning at 6 pm are locked in at 4 pm
      • ‘the hour beginning at 6 pm is known as the hour ending 19 (HE 19. this is because the hour ends at 7 pm, which is 19:00’
        • what the fuck does this mean
      • within two hours before delivery begins, quanities can change only for an acceptable operating reason, typically a chnage in the capability of the unit
    • within the delivery hour
      • SC (system controller) dispatches units up or down according to offers as demand changes
      • SMP is the offer price of the last dispatched unit in each minute; as demand changes, the marginal unit and SMP change
        • this will chnage substantially once alberta implements dispatch and pricing derived from optimization
      • units ramp towards instructed level
        • regulating reserve covers the small supply-demand imbalances that remain during ramping
      • time-weighted average: TRA of the SMPs is the hourly pool price: an SMP of 90/MWh for forty minutes yields a pool price of $80/MWh.
      • most consumers do not obersve the current price and do not respond to it: some large consumers do
    • within the following one to six months
      • financial settlement occurs for the delivery month: generators are paid, laqrge consumers and retailers pay
      • most consumers instead pay rates negotiated with their retailers, which do not vary hour to hour
      • market paritcipants are assessed for compliance with the rules and legislation
        • subject to investigation and enforcement
      • enforcement processes generally occur after the delivery hour, subject to statutory limitations periods
  • examples of market data:
    • monthly average pool prices
    • pool prices and temperature
      • what the fuck is meant by temperature
    • intermittency and the pool price over one week
    • next day: