Task Force Report MJohannis Draft Task Force Ris by MikeJenny


More Info
									Task Force 1-2
I   Objective/ Overview

The increasing penetration of variable generation resources makes it important to define ―best
practices‖ for quantifying the contribution of these resources to resource adequacy. The most
common resource adequacy metrics are Loss-of-Load Expectation (LOLE) and planning reserve
margin (PRM). This paper describes how various NERC regions count variable generation
toward capacity and energy adequacy. Because most regions are capacity-constraint the focus
will be on capacity adequacy. The goal of this paper is to define a compendium of ―best
practices‖ for evaluating variable generation’s contribution to resource adequacy.

Variable generation is intermittent in nature and non-dispatchable and is primarily comprised of
generation from wind and solar photo-voltaic (PV) resources. Wave energy may also fit this
definition, but proponents believe it is more predictable than other types of variable generation
and so might be included in the energy-limited category of resources. Variable generation’s
attributes will vary by geographic area and climatic regime, so it is entirely possible to have
wind, for example, contributing 60% of its installed capacity toward capacity adequacy in one
area and 0% in another area. In order to quantify this contribution, it is necessary to have a
sufficiently long data record to be able to evaluate the statistical attributes of variable generation,
and also whether there are any statistical relationships with other important parameters such as
load (i.e. via temperature).

NERC has a number of teams working to define all aspects of variable generation’s impact on
and contribution toward reliability. Quantifying variable generation’s contribution to resource
adequacy is within the scope of both the Resources Issues Subcommittee (RIS) and the
Integration of Variable Generation Task Force (IVGTF). For this reason, a joint RIS-IVGTF
team is drafting this paper.

RIS has proposed an ―operable capacity‖ metric to the NERC Planning Committee (PC) as a way
of quantifying the expected contribution of variable, energy-limited and machine-limited thermal
generation and dispatchable demand response resources to meeting loads and operating reserve
requirements over the peak load period in the PRM calculation. A big benefit of using the
operable capacity metric is that it allows comparability among resource types and transparency
as to the components of the PRM equation. The PC has provided positive feedback for the RIS
to proceed to develop this metric in coordination with the Reliability Metrics Working Group
(RMWG). The methods defined by this paper to calculate the contribution of variable generation
to the peak period PRM will allow the quantification of wind and solar PV operable capacity.
Please refer to graphic on the next page for a depiction of how the derates defining operable
capacity allow for comparability among the various resource types.
                                          Operable Capacity using Unforced Capacity (UCAP) Approach

Installed Thermal                                    Installed Energy-                             Installed Variable
    Capacity                                         Limited Capacity                                                                    Demand
                                                                                                        Capacity                        Response
(Gas-fired, Coal, Nuclear,                            (Hydro, Solar-Thermal,
                                                                                                       (Wind, Solar PV)
       Geothermal)                                          Biomass?)

                                                                                                                Reduced to reflect
               Mechanical Availability

                                                                                                           Probability of Availabilty
                Air Quality Limitations

                                                                  Reduced by Fuel
                                                                  Limitations under
                                                                 adverse conditons
                                                            & Mechanical Availability

                                                                                                                                           Reduced by expected
                                                                                                                                           availability & DR limits
         Reduced: Temperature derate

                                                                                                                     when Needed
                                                                                                          & Mechanical Availability
                                                                                        Task 1.2 Team

                                                         Operable                                        Operable                       Operable DR
 Operable Thermal                                      Energy-Limited                                    Variable                        Capacity
     Capacity                                            Capacity                                        Capacity
                                                                                             IVGTF-RIS Task
                                                                                                1.4 Team
                                                                                        Flexible Resources
                                                                                        to Firm & Integrate

II   Traditional Planning Techniques
     A LOLE & LOLP – Michael M.
     B Principles of ELCC – what is it, how does it work - Michael M.
        1 Calculation Methods
        2 Examples of Regions applying traditional techniques – Mary J.

           3                  REGIONAL EXAMPLE: Northwest Methodology for Calculating Wind
                              Generation’s contribution to Capacity Adequacy:

In April 2008, the Northwest Power and Conservation Council (NWPCC) adopted a voluntary
resource adequacy standard for the Northwest, which includes the states of Oregon, Washington,
Idaho and western Montana (Refer to link at http://www.nwcouncil.org/library/2008/2008-
07.htm ) Because of limited data availability, the contribution of wind generation to meeting the
capacity PRM metric was initially estimated at 15% of its installed capacity since this was the
amount of wind generation available in the Bonneville Power Administration (BPA) balancing
authority (BA) over the sustained peaking period during the July 2006 extreme heat wave event.
However, with increasing amounts of wind generation and longer periods of generation records,
it has become increasingly apparent that wind generation cannot be counted to be available at
15% of its installed capacity when needed most. Currently, the regional estimate for wind’s
contribution to capacity adequacy is 5%, which is likely to decrease further (See example
graphic below).

                           BPA Balancing Authority Area Load & Total Wind Generation
                                                                                                       Jan. 5-25, 2009







                                                                                                                                       BPA TOTAL WIND GENERATION

          2000                                                                                                                         BPA BALANCING AUTHORITY AREA LOAD






















                                                                                        Date/Time (5-min increments)

The methodology currently under development for calculating wind’s contribution to Northwest
resource adequacy involves the use of the NWPCC’s GENESYS Loss-of-Load Probability
(LOLP) model. When there were only 300 MW of wind in the region, wind was modeled as a
thermal resource in GENESYS. Now that wind resources are an order of magnitude higher with
penetrations of over 20% in some BAs, such as the BPA BA, the following method is under
development to evaluate wind’s contribution to resource adequacy:

     Monte Carlo picks of hydro generation, thermal forced outages and temperature years will
      be performed by the GENESYS model;
     This methodology assumes that hydro generation, thermal forced outages and annual hourly
      temperatures are independent variables, which collectively capture the majority of the
      power system uncertainty; it also assumes that wind generation and load are not
      independent variables because both are influenced by temperature; therefore, temperature is
      assumed to be the independent variable.
     The temperature-year pick allows for the calculation of regional load—there is a significant
      temperature-sensitive load component in both the summer and winter;
     The temperature-year pick also allows for the selection of a long-term, temperature
      correlated synthetic wind generation record; these records are being developed using the
      ―nearest neighbor‖ method, which results in hourly synthetic wind generation data that is
      correlated to historical temperatures and retains the statistical attributes of wind in terms of
      hourly variability, plant factors, etc.
     Theoretically, LOLP studies with and without wind generation should allow for calculation
      of the contribution of wind capacity toward adequacy under peak load conditions, i.e. the
      operable capacity of wind.
     In addition, an examination of Loss of Load (LOL) events, which are generally comprised
      of periods of high loads (due to cold snaps or heat waves) and low hydro generation will
      allow an examination of wind’s contribution to meeting load (plus operating reserves)
      under these events.

III Application of traditional techniques on VG – Gary J & Michael M.
    A How is VG different from traditional generation
    B How do the VG differences influence the calculation?

IV Data needs. Discussion of correlation wind, hydro, solar, etc.

V Approximation methods (such as capacity factor over suitably-defined peak periods).
     Michael has some material as a starting point.

VI What are people doing for VG capacity credit today - Mahendra; Mary will provide
     information about what RAS is doing. Michael has some older material that can serve as a
     starting point.

BPA Example: As of November 2009, the BPA BA has 2,253 MW of installed wind capacity
for a net internal load of about 10,500 MW. This results in an over 20% wind penetration. BPA
has developed the following approach for calculating wind capacity credit by evaluating the
contribution of adverse wind generation toward capacity adequacy in a similar manner as the
contribution of adverse hydro generation toward adequacy is evaluated.

The capacity needs of BPA change depending on the month. There are times of the year when
capacity is of particular concern. The months where we have the highest loads and the least
flexibility in the hydro system are in the winter and late summer. In particular the months of
December, January, and February are of concern for the winter season and the months of July,
August, and September are of concern for the late summer months. The fifth percentile of the
hourly capacity factor for each month of the year is given in Table 1 and the fifteenth percentile
is given in Table 2. These percentiles are based on the observed hourly output of the BPA
integrated wind fleet from 2003 through the end of 2008.

 Table 1: Fifth percentile of the observed BPA integrated wind fleet hourly capacity factors by

              Month                                                     Capacity Factor
              January                                                            0.00%
              February                                                           0.00%
              March                                                              0.36%
             April                                                                0.18%
             May                                                                  0.32%
             June                                                                 0.48%
             July                                                                 0.47%
             August                                                               0.22%
             September                                                            0.08%
             October                                                              0.00%
             November                                                             0.00%
             December                                                             0.00%

 Table 2: Fifteenth percentile of the observed BPA integrated wind fleet hourly capacity factors
                                            by month.

             Month                                                      Capacity Factor
             January                                                             0.04%
             February                                                            0.53%
             March                                                               2.50%
             April                                                               2.47%
             May                                                                 2.20%
             June                                                                2.69%
             July                                                                2.51%
             August                                                              1.41%
             September                                                           1.22%
             October                                                             0.61%
             November                                                            0.41%
             December                                                            0.18%

From both of these tables we can see that there is a minimal contribution toward capacity during
the winter months. Even at the fifteenth percentile this could be treated effectively as zero
contribution of wind toward capacity. At the fifth percentile the summer capacity contribution is
also effectively zero, only at the fifteenth percentile during the summer is there a possibility of a
non-zero contribution to capacity of wind. BPA has decided to use a zero value for wind
capacity for both winter and summer based on this approach.

VII Recommendations
    A Suggestions for GADS data collection
    B Recommendation of method(s) to PC

Draft report/powerpoint by end of February for March PC meeting.

Final report due mid/late May in time for June PC meeting.

Materials provided by Michael Milligan (we can borrow any material if appropriate):

http://www.nrel.gov/docs/fy08osti/43433.pdf - note the table on page 24. Things have changed,
but this may be helpful as a starting point.


To top