Wind energy delivery issues [transmission planning and competitive electricity market operation].pdf

T THE RAPID INCREASE IN WIND GENERATION IN RECENT YEARS

has raised the visibility of an issue that has been troubling power grid plan-

ners and operators for years: how to deal with large amounts of intermittent

generation resources connected to the grid. Grid planning and operating

practices, including the structure of the electric power market, are largely

based on dispatchable generating resources (i.e., generators capable of pro-

ducing power up to their full rating whenever the system operator schedules

them to do so). A typical system-operating scheme follows this sequence:

Day-ahead forecast: Market participants forecast system load for each

hour of the following day. This is a sophisticated process involving

historical information, weather forecasts, and time of day.

Day-ahead market: Generators and load-serving entities bid for

producing and purchasing energy and operating reserves.

3) Unit commitment: System operator schedules an appropriate mix

of generating resources to serve the load recognizing factors such

as bid prices for energy, generator start-up and maneuvering

constraints, and transmission congestion constraints.

4) Real-time operation: System operator adjusts generating

resources to match actual system load in real time during

the day of operation.

5) Market settlement: Actual power generated and con-

sumed is logged, and imbalances from

scheduled values are financially settled,

following a prescribed set of market rules.

november/december 2005

IEEE power & energy magazine

Wind generators introduce new challenges due to their

intermittent nature. The amount of power a wind generator

can produce depends on the wind conditions at the time.

Although wind generator output can be forecast a day in

advance, forecast errors of 20–50% are not uncommon.

These characteristics of wind generation increase the levels

of variability and uncertainty in power grid operations.

Transmission system planners are faced with a related set

of challenges. Renewable portfolio standards (RPSs) that set

minimum requirements for renewable energy are being

adopted worldwide. In the United States, production tax cred-

its are (sometimes) available to encourage development of

new generating resources. As a result, new wind generation

projects are numerous and increasing. Wind capacity in the

United States is expected to exceed 9,000 MW in 2005, up

from about 6,800 MW in 2004. New York State presently has

about 50 MW of installed wind generation and over 4,000

MW of new projects in the queue—and this is not unique.

Many of the best wind resources are remote from load

centers or existing transmission corridors. In most areas with

deregulated power markets, existing planning practices do

not look ahead towards expanding the transmission grid to

serve such resources. Individual wind projects cannot afford

to pay for such major transmission expansion on their own,

so the wind resources may remain stranded and undeveloped.

What can be done to address these issues? Should regional

transmission operators take on the responsibility to install

transmission in anticipation of new generation resources?

Should traditional system operating practices be changed to

accommodate intermittent generating resources without com-

promising grid reliability? Should power market rules be

changed to accommodate the nondispatchable nature of wind

generation, and could this be done while maintaining a fair

and competitive market?

eration Interconnection Queue. Class 4 (good) or higher wind

development areas were the primary location of the wind gen-

eration in the queue last year, but Class 2 (marginal) queue

entries have been occurring more frequently this year.

The Midwest ISO planning process is an open, collabora-

tive process with the transmission owners and other stake-

holders. Inputs are obtained from a wide range of

stakeholders from the Midwest ISO advisory groups and

from various meetings. That input is then studied using multi-

ple methods to resolve the problems of providing adequate

transmission for all the transmission system requirements,

including wind generation.

Bottom-up transmission studies compile the individual

requirements for transmission into one system. That system is

then tested and adjusted to meet planning criteria. Examples

include the reliability portion of the MTEP (which assembles

all the transmission owners plans into one master plan), the

interconnection queue sequential studies under FERC

processes, and the transmission requirements to serve load in

the Midwest ISO. One such bottom-up study, the Buffalo

Ridge study, is determining the short-term need for transmis-

sion to serve wind generation in the particularly congested

area of southwestern Minnesota. Multiple study groups have

studied problems and recommended local solutions. The

local solutions are incorporated and tested, and reliability vio-

lations are resolved in the MTEP process.

Top-down transmission studies are based on generation

scenarios that are formulated in an open, collaborative

process by stakeholders for some specified future period.

Top-down processes only provide information and do not

determine the transmission that must be built. An example of

such a study is the MTEP 03. It included a study of 10,000

MW of wind generation to determine the transmission

table 1. Potential wind generation in MISO region.

Midwest Independent System Operator

Transmission Planning Process and the

Implications for Wind Generation

The potential for wind generation within the Midwest Indepen-

dent System Operator ( ISO) footprint is significant. As Table 1

shows, there is more than 500,000 MW of potential wind gen-

eration capacity in the Midwest ISO and neighboring areas. A

10% renewable energy objective for the entire Midwest ISO

footprint today would require about 19,000 MW of wind gen-

eration, but that represents only 4% of the total potential wind

generation capacity available. The Midwest ISO has 5,800

MW of wind capacity in the Generation Interconnection

Queue as of February 2005, and an additional 5,000 MW of

wind generation under study for the Midwest ISO transmis-

sion expansion plan (MTEP) exploratory studies. The current

level of installed wind network resource generation on the

Midwest ISO system is 860 MW.

Most potential wind generation is in remote locations, as

shown on the map in Figure 1. Wind generation constitutes

65% of the total number of requests in the Midwest ISO Gen-

Wind Power (MW)

Existing MW 1

Total Potential MW 2

State

Illinois

6,980

Iowa

471

62,900

Minnesota

563

75,000

Nebraska

99,100

North Dakota

138,400

South Dakota

117,200

Wisconsin

6,440

Total

1,261

506,020

Notes:

1 Nameplate MW (American Wind Energy Association,

Jan. 2004, http://www.awea.org/)

2 Average MW, circa 33% of nameplate capacity (from “An

Assessment of Windy Land Area and Wind Energy Potential,”

Pacific Northwest Laboratory, 1991) .

Source: Wind on the wires presentation on net environmental

impacts of transmission systems in the Midwest.

IEEE power & energy magazine

november/december 2005

needed to deliver the wind energy without undue transmis-

sion constraints. The interconnection queue was used as an

input, but the generation scenario was also formed by stake-

holder input. MTEP 05 further refined the exploratory plans

for combinations of wind generation and coal in the Dakotas

and Minnesota and additional plans for northern Iowa, south-

ern Minnesota, and Wisconsin. One use of the information

from the top-down processes is to provide realistic examples

for regulatory and legislative processes. Generators may use

the information to plan their interconnection queue entries.

CAPX is an integrated top-down capital expenditure study

of generation options for Minnesota and the surrounding

states for horizon year 2020. CAPX also addresses the trans-

mission necessary to serve those options. The 10% renewable

energy objective for Minnesota is modeled in the studies. The

Midwest ISO transmission owners have worked with the

State of Minnesota on the study and on deriving the legisla-

tive requirements to implement and recover costs. The Mid-

west ISO is participating in the economic studies for CAPX,

which should result in a recommended blueprint for future

options for transmission and generation development for the

area. The results of CAPX will then be included in the MTEP

process when appropriate.

In summary, the study process and associated regulato-

ry and legislative process are still evolving. Progress is

being made and methods are being formulated and exer-

cised to resolve the issue of providing transmission on a

regional basis for generation that includes wind in the

Midwest ISO footprint. Similar efforts are underway in

other ISOs across the country.

Integration of Wind Generation into the

California ISO Markets and Operations

California is a leader in the United States in the develop-

ment of renewable resources including wind, solar, geother-

mal, biomass, and small hydro generation resources. In

September 2002, the state passed legislation (SB 1078) that

created the California Renewables Portfolio Standard

(RPS). This law requires the investor-owned utilities (IOUs)

to increase their procurement of renewable energy to 20%,

based on the total energy they deliver to customers by 2017.

The new energy action plan for the state accelerated this

goal to 20% by 2010.

Wind generation will supply a major part of the renewable

energy required to meet the RPS goal. Wind is forecasted

to increase from 4,000 GWh of energy in 2004 to over

15,000 GWh by 2010. The installed capacity is forecasted to

increase from 2,100 MW in 2004 to 7,500 MW in 2010. This

forecasted increase in wind generation requires solutions to a

number of issues:

✔

market integration

✔

real-time grid operations

✔

calculation of the capacity value of wind generation

✔

solutions for environmental impacts

United States Annual Average Wind Power

MISO Generation

Wind Power

Queue Entry

Locations

Power Classification

1 Poor

2 Marginal

3 Fair

4 Good

5 Excellent

6 Outstanding

figure 1. Wind resource map showing locations of planned wind projects in Midwest ISO region. (The map is from the

U.S. Department of Energy and the National Renewable Energy Laboratory; the queue overlay is from Midwest ISO.)

november/december 2005

IEEE power & energy magazine

✔

interconnection standards

will be needed. Other areas requiring further exploration

include new concepts for the automatic generation control

and dispatch of controllable loads to assist with mitigating

the impact of wind generation variability. The California ISO

has established a working group to address these and other

operational issues. The group includes participants from the

California ISO, the wind generators, the utilities, and the Cal-

ifornia Energy Commission. The recommended solutions will

be published by December 2005.

California’s formula for calculating capacity value of

wind generation is based upon three years of wind energy

production data during the hours of noon to 6 p.m. for the

months of May through September. The blue bars in

Figure 2 show the average hourly wind energy production for

2004. The yellow bars show the average hourly energy pro-

duction during the peak hours for these five months as well as

a capacity percentage based on a total of 2,046 MW of avail-

able capacity. The amount of wind generation energy to meet

summer peak loads declines significantly after June and, in

fact, was less than 300 MW, or 15%, at the peak hours on the

hottest days of August and September 2004.

Work is continuing on other issues such as interconnection

standards and transmission planning requirements for wind

✔

planning and construction of new transmission

transmission availability and utilization.

California has addressed the first problem: the integration

of wind generation into the energy markets. The California

ISOs Participating Intermittent Resources Program (PIRP)

went into operation in August 2004. The PIRP lowers the

risks for wind generators of bidding into forward energy mar-

kets without incurring 10-min imbalance energy charges.

Wind generators must schedule their energy in the hour-

ahead market by using an advanced forecasting service, and

this forecast becomes their deemed delivered schedule. Devi-

ations between actual energy delivery and the scheduled

amount are multiplied by the hourly price, and the total dollar

amounts are collected in an account for settlement at the end

of the month. An unbiased forecast of hourly energy produc-

tion should result in a relatively small net energy deviation

over the entire month.

There are currently ten participants with a total of 450 MW

of wind generation capacity enrolled in the PIRP program.

These generators produced over 530,000 MWh of energy in

2004. However, this only represents 20% of the total wind

generation available. The other 80% is currently covered by

qualifying facility (QF) contracts with

local utilities, and the utilities have

responsibility for forecasting and

scheduling the energy. When the Cali-

fornia ISO implements uninstructed

deviation penalty charges, the utilities

will have a greater incentive to sched-

ule the wind generation energy

through the PIRP program.

Although the existing 2,100 MW of

wind generation in California has not

resulted in serious operational issues, it

does have a noticeable impact on oper-

ations. The most serious of these prob-

lems is over-generation during the

night. Wind energy production is high

during the late spring months, the

same time hydro generation is at its

peak due to the melting snow in the

mountains. The load is low during this

period, and the goal is to have other

generation off line or ramped down. Ultimately, some wind

generation may have to be curtailed during this period to mit-

igate the over-generation condition. There is also a need for

new procedures and protocols for controlling large ramps

both up and down during major storms that cause high wind

variability.

Data from current operations is being used to assess future

operational issues when the installed wind generation

capacity increases to 7,500 MW or more. New methods are

needed for calculating, on a seasonal and day-to-day basis,

the amount of regulation and load-following resources that

✔

1,000

24 h Avg. Production

Peak Hours

800

600

400

200

May

June

July

Aug.

Sept.

figure 2. Average hourly wind energy production during 2004.

generators. The Western Electricity Coordinating Council

(WECC) has proposed a new low-voltage ride-through stan-

dard that is less restrictive than the standard FERC adopted.

The transmission expansion plan for the Tehachapi region in

Southern California will be the test for how to plan and finance

the transmission network in order to move 4,000 MW of new

wind generation to the California load centers. Transmission

capacity upgrades take a lot longer to plan and construct than

the corresponding time required to build new wind generation

plants. If a transmission company builds a major transmission

line to a wind generation area with the expectation that new

IEEE power & energy magazine

november/december 2005

wind generating facilities will be built there, can the transmis-

sion company be assured they will earn an acceptable rate of

return on the investment? This question has been submitted to

FERC, and California is awaiting a FERC ruling on a proposal

for new transmission for the Tehachapi area.

scenario. The base approach involves using day-ahead wind

generation forecasts for the unit commitment process, and

adjusting the hydro generation after scheduling the wind out-

put. Operating cost impacts, based on the 2001 historical

hourly load and wind profiles, are summarized in the first

column of Table 2. These are impacts for New York ISO

(NYISO) only and do not include additional savings in New

England and PJM. The total amount of wind energy generat-

ed in the 2001 simulation was approximately 8,900 GWh.

Therefore, the NYISO variable cost reduction in US$/MWh

was calculated to be US$350 million/8,900 GWhr

Lessons Learned from NY State

Study of 10% Wind Penetration

March 2005 marked the release of the final report on “The

Effects of Integrating Wind Power on Transmission System

Planning, Reliability and Operations” of the New York State

power grid. This 18-month study by GE Energy Consulting

examined the addition of 3,300 MW of wind generation

(approximately 10% of the system peak load). The overall

conclusion was that the NY State Bulk Power System

(NYSBPS) could readily accommodate this level of wind

generation with only minor adjustments to the existing plan-

ning, operation, and reliability practices. Some of the key

impacts on system operations and effective capacity are dis-

cussed below.

US$39/MWh. The simulation results also indicated a

US$1.80/MWh average reduction in spot price in New York.

The operating costs depend on how the wind resources are

treated in the day-ahead unit commitment process. If wind

generation forecasts are not used for unit commitment, then

too many units are committed and efficiency of operation suf-

fers. The operating costs for this situation are summarized in

the second column of Table 2. In this case, unit commitment

was performed as if no wind generation was expected, and

wind energy just shows up in the real-time energy market.

The results indicate that energy consumers benefit from

greater load payment reductions, but nonwind generators suf-

fer due to the inefficient operation of committed units. In

Table 2, the third column compares the two cases and shows

System Operating Costs

GE’s Multiarea Production Simulation (MAPS) program was

used to simulate the hourly operation of the NYSBPS for sev-

eral years, with and without wind generation per the study

140

120

Load

Load - Wind

100

−

2,200

−

1,800

−

1,400

−

1,000

−

600

−

200

600

1,000

1,400

1,800

2,200

2,600

figure 3. Statewide hourly variability for January 2001.

november/december 2005

IEEE power & energy magazine

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: