| Description: | A market that can supply and/or demand a Commodity |
See also Market Property Reference for a detailed list of properties for this class of object.
The Market class is a versatile tool in both market modeling (where the objective function is maximization of net benefit - essentially minimization of system cost) and portfolio optimization (where the objective function is profit maximization).
In market modeling, a Market object represents a source and/or sink for a commodity e.g. electricity, gas, etc (see below for a complete list). Transactions in the market become part of the system cost minimization objective function i.e. Purchases from the market are 'costs' and Sales to the Market are 'benefits' (or 'negative costs'). The properties Max Purchases and Max Sales determine the volume of purchases/sales allowed in the Market. Thus a Market can be either or both a source and a sink. The price paid for purchases and/or received for sales is set by the Price property. This can be either a fixed price or defined as in multi-band pairs of Price and Quantity with Base Quantity to define a market supply/demand curve for a market with depth i.e. where the price varies depending on the amount purchased/sold. An alternative to defining a supply/demand curve is to set a fixed Demand for the given commodity and optionally Shortage Price and Surplus Price. This creates a simple 'vertical' demand function with price caps and is convenient for modeling markets that do not have demand bidding.
In portfolio optimization, the primary input property is the Market Price (which, like most inputs, can be a deterministic or stochastic variable). The objective function becomes maximization of revenue less production cost i.e. maximize profit from trades in the market.
Market objects can be buyers or suppliers of the following:
Markets can also be used in a passive sense to 'mark-to-market' Generator Generation or Reserve Provision i.e. to price the production.
Any number of markets can be defined in a simulation and the simulator co-optimize sales and/or purchases to/from all markets simultaneously. Trades in markets can be combined seamlessly with 'local' load, ancillary services, etc.
Markets can represent either external markets in which trades are made, or they can represent options or transactions such as physical contracts for generation, load obligations, fuel nominations that are exercisable over some time period e.g. a daily exercisable option.
Markets can be either forward or real-time as described by the Is Forward property. Forward markets are optimized first in a two-stage optimization process. Once the optimal level of forward market sales and purchases has been determined these levels are fixed while the real-time markets are solved. This two-stage optimization ensures that there is no "pure trading" between forward and real-time markets. The two-stage solve is invoked automatically hence there are both forward and real-time markets in the model.
Each Market object provides collections for connecting the external market to the other simulation elements. How this connection works depends on the type of market being modeled.
| Type |
Description |
| Commodity | A Market can buy/sell any Commodity via the Markets membership. |
| Electricity | An energy market is connected to one or more Nodes. Trades to and from the energy market are made at these nodes. The megawatt trades act like pseudo loads (for sales) or generators (for purchases). Note that the nodal price will still reflect the marginal cost of production at the node i.e. the external market prices will not always set the prices at these buses. |
| Capacity | A capacity market is connected to a set of generators. Those generators receive revenue from the capacity as a function of their availability. |
| Heat | A market for heat is connected to one or more Generators that can provide waste heat e.g. for district heating requirements. The conversion rate between the heat input of the generator and the waste heat provided to the market is defined by the property. |
| Fuel | A market can be connected to one or more Fuel objects. This feature is designed to model arbitrage between a Fuel Contract and a spot market with the Market object representing the supply of spot market fuel. |
| Emission | Emissions can be bought and sold to/from a Market by connecting the Market to one or more Emission objects. Rather than trading the emissions, the sales and purchases into another market e.g. an energy market can contribute to or abate emissions if you define either the Emission Markets Sales Coefficient or Purchases Coefficient. |
| Reserve | Each ancillary services market must be 'connected' to a reserve. The reserve object defines the set of generators that provide the service and their capabilities. Trades can be made in either direction i.e. sales to the external market or purchases to meet local reserve requirements. Note that the reported reserve price will still reflect the marginal cost of reserve provision i.e. the external market prices will not always set the prices of reserve, but all trades will be valued according to the market prices. |
| Gas | Natural gas can be bought from or sold a Market by connecting the Market to one or more Gas Node objects. The Market then acts like an export (sales) or import (purchases) at the gas node. |
As well as representing complete external markets, Market objects can be used to represent specific exercisable options for either supply or demand. Thus a portfolio optimization could include many Market objects representing available options that are to be optimized exercisable over various time frames.
Options for buying from a market are set up using Max Purchases and one of Buy Block and-or Buy Block Day/Week/Month/Year, and Buy Unit.
Options for selling to a market are set up using Max Sales and one of Sell Block and-or Sell Block Day/Week/Month/Year, and Sell Unit
The Block properties (Buy Block, Sell Block and the period derivatives) determine the period over which the option is exercised (i.e. period, day, week, month, year). The Buy Unit ands Sell Unit properties determine the discreet size of the option. For example a buy block of 25 MW allows the option to be exercised in 25 MW increments up to Max Purchases.
Markets can be settled using marginal pricing or volume-weighted pricing- the default being marginal pricing. This is controlled by the Is Marginal property.
The volume and value of trades made to/from the market is reported in the properties: Sales, Purchases, Net Sales, Net Purchases, Revenue, Cost, Net Revenue. The volume-weighted price for sales/purchases are reported as Price Received and Price Paid respectively.
To flow these reported trades and revenues/costs back to the portfolio owner you need to add the Company or companies that own the trades in the Market Companies collection. You can share the reported volumes and revenues/cost using the property of that membership.
Note that generators will continue to receive income from energy sales into their transmission node and those incomes will flow back to the owning companies as reported pool revenue. Because some or all of this generation is sold into energy markets there will be some double-counting of revenues. There are couple of ways of dealing with this, you can either:
The market class is a very flexible model and with some of the markets it is possible to inadvertently create an unbounded objective function. This is reported by the solver as a dual infeasibility (if the dual problem is infeasible this means the primal problem's objective is unbounded). This situation can happen where your data implies that infinite sales can be made into a market with a non-negative price. You should correct this problem e.g. by setting a cap on sales using Max Sales.