Generator Class
| Description: | generating unit, or collection of like generating units |
| Detail: |
See also Generator Property Reference for a detailed list of properties for this class of object.
Contents
- Generator Type
- Categories
- Network Connection and Floating Generators
- Price Received
- Capacity and Outages
- Production Cost
- Unit Commitment
- Stochastic Unit Commitment
- Ramping
- Auxiliary Use
- Emissions
- Constraints
- Capacity Expansion
- Anti-Generators (Generators as Loads)
1. Generator Type
All types of generation resource e.g. thermal, nuclear, hydro, pump storage, RES, etc. are modeled with the Generator class. There is no need to specifically identify the type of a generator; the simulator infers the type of generator from the data and relationships that are defined on it. Regardless of the type of generator, you must define both Units and Max Capacity. Note that the default value of Units is zero, and hence you must define it for the generator to 'exist'.
1.1. Thermal Generation
A thermal generator uses one or more fuels. The efficiency of generation is defined using a heat rate function. You can chose to define the fuel cost using the Fuel Price property directly, or indirectly by adding one or more fuels to the generator's Fuels collection, and setting the Price property of the fuel(s). Thermal generators define a normal operating range using the property Min Stable Level.
Example (using metric units)
| Property | Value | Units | Date From | Date To |
|---|---|---|---|---|
| Units | 1 | - | 1/01/2001 | |
| Max Capacity | 300 | MW | ||
| Min Stable Level | 140 | MW | ||
| Heat Rate | 11.5 | GJ/MWh | ||
| Fuel Price | 4.5 | $/GJ |
Example (using imperial US units)
| Property | Value | Units | Date From | Date To |
|---|---|---|---|---|
| Heat Rate | 11500 | BTU/kWh | ||
| Fuel Price | 4.5 | $/MMBTU |
NOTE: To switch between metric and imperial units for data entry use the Settings form.
The operation of combined-cycle generators can be optimized in detail as described in the Combined Cycle Plant article.
Combined Heat and Power (CHP) plant are modeled using special input properties, as described in the Combined Heat and Power article.
1.2. Nuclear Generation
Nuclear generators are often run to a strict regime, with a high minimum operating level, which is defined with Min Stable Level. To define a fixed operating regime, use the Fixed Load or the Min Load properties.
Example
| Property | Value | Units | Date From | Date To | Pattern |
|---|---|---|---|---|---|
| Units | 1 | - | |||
| Max Capacity | 1500 | MW | |||
| Min Stable Level | 990 | MW | |||
| Min Load | 1080 | MW | W2-6 | ||
| Min Load | 990 | MW | W1,7 |
1.3. Hydro Generation
Hydro generators can be modelled in a number of different ways. The simplest approach is to define energy constraints to approximate the availability of water (see Max Energy, and Max Capacity Factor), and a profile of minimum operating levels to represent run-of-river generation (see Min Load).
Example
| Property | Value | Units | Date From | Date To | Pattern |
|---|---|---|---|---|---|
| Units | 1 | - | |||
| Max Capacity | 60 | MW | |||
| Min Load | 5 | MW | M1-4,10-12 | ||
| Min Load | 15 | MW | M5-9 | ||
| Max Energy MONTH | 15 | GWh | M1-4,10-12 | ||
| Max Energy MONTH | 27 | GWh | M5-9 |
Alternatively if you have a pre-defined profile of hydro generation you can enter this using the Max Load property.
A hydro generator with storage must have a Head Storage defined. This membership connects the generator to the storage model. Generation will then subtract from the storage according to the defined Efficiency function. To create a cascade of storage, connect the Generator to a Tail Storage. Spillways, and canals are modeled with Waterway objects. See the article Cascading Hydro Networks for more details.
1.4. Pumped Storage
Pumped storage plant must have a Head Storage and a Tail Storage defined. In addition you must define the properties Pump Efficiency and Pump Load. You may also define Pump Units and Min Pump Load. Pump storage is usually modeled as a closed system, i.e. no energy flows into or out of the head or tail storage other than from generating or pumping. See the article Pumped Storage for more details.
Example
| Property | Value | Units |
|---|---|---|
| Units | 1 | - |
| Max Capacity | 240 | MW |
| Pump Load | 240 | MW |
| Pump Efficiency | 70 | % |
1.5. Renewable Energy Sources and Energy Storage
Renewable Energy Sources (RES) and battery energy storage systems (BESS) can be modeled to a high level of detail. See the article Renewable Energy Sources.
1.6. Synchronous Condenser
A synchronous condenser is a generator connected and synchronized to the power grid but not generating electricity. A unit running in this mode can provide spinning reserve or inertia. You can indicate that a generator can run in this mode by setting the Sync Cond Units input property. A synchronous condenser can draw a small amount of load from the grid according to the Sync Cond Load property.
The following example illustrates a generator that can switch between generating and synchronous condenser modes.
| Collection |
Property | Value | Units |
|---|---|---|---|
| Generator | Units | 1 | - |
| Generator | Max Capacity | 20 | MW |
| Generator | Sync Cond Units | 1 | - |
| Generator | Sync Cond Load | 0.1 | MW |
| Generator | Sync Cond VO&M Charge | 10 | $/MW |
The following example illustrates a generator that acts only as a synchronous condenser to provide inertia. Note the use of Generating Units = 0 to indicate that the Units are not allowed to generate. In this case each unit operating in synchronous condenser mode provides 20 × 7 = 140 MW⋅s inertia.
| Collection |
Property | Value | Units |
|---|---|---|---|
| Generator | Units | 1 | - |
| Generator | Max Capacity | 20 | MW |
| Generator | Generating Units | 0 | - |
| Generator | Sync Cond Units | 1 | - |
| Generator | Sync Cond Load | 0.1 | MW |
| Generator | Sync Cond VO&M Charge | 10 | $/MW |
| Generator | Inertia Constant | 7 | MW⋅s |
2. Categories
As with any class, generators can be organized into categories. The categorization of generators does not affect the way they are modeled but it does help when reviewing and editing input and solution data. Categorizing your generators is highly recommended.
3. Network Connection and Floating Generators
A generator must be connected to the transmission network by adding one or more nodes to the generator's Nodes collection. Note that you may apportion the generation across multiple buses if required using the Generation Participation Factor property. For pumped storage the load can be distributed to one or more nodes with the Load Participation Factor.
The 'wildcard' Nodes* collection is used to create a copy of the Generator at each Node. This is useful in LT Plan for creating 'floating' generators i.e. projects whose location is to be optimized along with the expansion decision. It is also useful in modelling small-scale distributed generation such as batteries or demand-side response. See the Nodes* topic for more details.
4. Price Received and Net Profit
The Price Received for Generation is controlled by the attribute Generator Settlement Model. The default is for generators to receive the weighted-average Price at the nodes they connect to but other options are available from this attribute. Alternatively you can mark-to-market by adding a Market object to the Mark-to-Markets collection of the generator.
A Generator object can Formulate Risk in a risk-constrained simulation to bounds its Net Profit within an Acceptable Risk. Alternatively, the Generator can participate in a Company which can Formulate Risk.
5. Capacity and Outages
Max Capacity and Units together define the installed capacity of the generating station. In multi-unit generators the units are assumed to have identical characteristics.
The units and/or max capacity properties should not be used to take units in and out of service for planned maintenance or mothballing. To model a case in which one or more units are out of service for periods of time, use the Units Out property with appropriate dates or patterns.
Examples
| Property | Value | Units | Date From | Date To |
|---|---|---|---|---|
| Units | 1 | - | 1/01/2001 | 1/01/2030 |
| Max Capacity | 150 | MW | ||
| Units | 4 | - | ||
| Units | - | 5/01/2010 | ||
| Units Out | 1 | - | 8/01/2004 | 9/01/2004 |
You can also model random unit outages, called forced outages by defining Forced Outage Rate. Additionally the placement of maintenance outages can be optimized - for a discussion of both random and planned outage modelling see the article Maintenance Scheduling.
6. Production Cost
Production cost is represented by offers to the market (also called 'bids'). Offers can be entered directly using Offer Quantity and Offer Price. But if user-defined offers are not defined the simulator creates a representation of the unit production cost function using the Fan Approximation method. This automatic method has a similar effect to defining a multi-band offer based on incremental cost, though it also captures the no-load cost.
Production cost is defined using a heat rate function, fuel price, and variable operations and maintenance cost.
An alternative to defining detailed offers is to define the 'economic range' (with Economic Maximum and/or Economic Minimum). This might represent the 'normal' operating range with generation outside of that range being reserved for ancillary services or times of shortage or over-generation.
7. Unit Commitment
The simulator optimizes generator unit commitment. The key properties relating to on/off decisions are Min Stable Level, Start Cost, and Shutdown Cost. You can also model more complex restrictions such as Min Up Time, Min Down Time
By default it is assumed that generating units can run up and down between zero and Min Stable Level instantaneously, with ramping inside the normal operating range restricted by Max Ramp Up and Max Ramp Down. However you can limit the ramping during start up and shutdown by using Run Up Rate (for a constant run up rate for each cooling state) or start profile (for a variable run up rate), and likewise Run Down Rate or shutdown profile for running down.
For operational modelling then initial conditions are defined using the properties Initial Generation, Initial Units Generating, Initial Hours Up, and Initial Hours Down.
Other constraints that affect unit commitment are Fixed Load, Min Load, and Rough Running Ranges.
Unit commitment can be hardwired for all or a selection of periods using the Must-run Units, and Commit properties.
May other constraints can be defined with properties or via the Constraint class including:
- Number of units started in a given period (day, week, month, year or other) using the Max Starts property and period variants.
- Number of hours of operation of a multi-unit station using the Constraints Hours of Operation Coefficient
There are a number of solution methods available for the unit commitment problem. The default setting relaxes any integer constraints and provides a simple Linear Relaxation. You should change the Production setting Unit Commitment Optimality to enforce integer constraints. The similarly named option Unit Commitment Optimality setting is also available on the Generator objects so that the solution method for unit commitment can be customized.
The trade off between solution quality and performance of the integer unit commitment can be controlled with the following settings:
- Generator Unit Commitment Aggregation
- Generator Unit Commitment Period
- Production settings related to heat rate detail, run up/down and start costs
- Performance settings for the MIP solver especially the MIP Relative Gap
8. Stochastic Unit Commitment
In the context of real-time operation of the generating system, unit commitment decisions must be made ahead of time. This is particularly the case for larger and/or slow start units that may require a notice period before they can begin a start up. Thus if we are optimizing the unit commitment of the system 'now' there will be a window of time over which the decisions we make to commit or not commit certain generators are locked in i.e. where there is no recourse available. This constraint, called the "non-anticipativity" constraint (because it is a constraint on the length of time we can react to unanticipated events) is modeled in the simulator using the special properties:
and for pumped storage pump commitment decisions:
These properties together define the penalty for "perfect foresight", with -1 indicating infinity, and the window of time over which is constraint applies.
Optimizing unit commitment with non-anticipativity constraints requires:
- Your simulation is defined with more than one sample of Variable values and/or outage patterns - see the Stochastic topic; and
- The simulation phase is set for stochastic optimization.
In ST Schedule you would set Stochastic Method = "Scenariowise-decomposition" (which is the method used to solve the stochastic optimization problem).
For all generators with non-anticipativity constraints defined the unit commitment decision over the defined time frame of the constraint will be optimized such that a single optimal commitment is made - optimal with respect to the uncertainty. This single optimal decision path will necessarily be more expensive than the solution of "Independent Samples" (perfect foresight) but is robust against the defined volatilities.
9. Ramping
The rate at which a generating unit changes load level is called the ramp rate. A unit's ramp rate is limited by its physical characteristics. Generally these limits imply that a unit cannot ramp through its entire operating range in any one dispatch interval. This limitation becomes more apparent as the length of interval decreases e.g. in a 5-minute market units can ramp only 1/12th as much per interval as an hourly market. The properties Max Ramp Up and Max Ramp Down set limits on the ramping capabilities. When modelling ancillary services using Reserve object it can be important to account for the combined effect of ramping and reserve provision. See the Response Ratio property for details.
10. Auxiliary Use
Generators use some energy 'in house'. This use is defined using the auxiliary use properties described in the property page Auxiliary Use.
11. Emissions
Emissions from production are modeled using the Emission class by associating the Emission with the Generators and/or Fuels. For more details, see Setting Emission Production Using Load Points and Bands.
Similarly, emissions abatement is modeled using the Abatement class by associating the Abatement with the Generators and Emissions.
12. Constraints
Certain common generation constraints are included as properties on the Generator object:
- Energy limits with Max Energy
- Capacity factor limits with Max Capacity Factor and Min Capacity Factor
Going beyond this capability, various aspects of generator dispatch and commitment can be included in generic constraints. These might be used to model fuel limits, emission limits, or transmission nomograms for example. Generic constraints can be defined for every period of the simulation, or over aggregate quantities e.g. daily, weekly, monthly, etc.
13. Capacity Expansion
The timing of new generation builds and retirements can be optimized using the LT Plan algorithm.
14. Anti-Generators (Generators as Loads)
The property Generation Participation Factor that defines the injection to the network is allowed to be negative. In this case the Generator is treated as a load by the Node. This can be a useful way to model very complex loads especially in LT Plan.