WIND GENERATION OPERATIONS
CHALLENGES AND EXPECTATIONS
OPERATIONS PLANNING STAFF
ERCOT
PUBLIC 2
Objectives• List two challenges in forecasting wind in Texas.
• List three actions a Wind facility experiencing icing issues should take.
• Given an example of telemetry being received from a renewable
resource, determine if it is an issue and whether the QSE representing
this resource should be contacted to rectify the issue.
PUBLIC 3
Wind Generation and Wind
Forecasting in ERCOT
PUBLIC 4
816 977 1,173 1,385 1,8542,875
4,785
8,0058,916 9,400 9,604
10,40711,065
12,470
15,76416,631 16,631 16,631
2,872
6,086 6,455248
2,291
3,952
116
816 977 1,173 1,3851,854
2,875
4,785
8,005
8,9169,400
9,60410,407
11,065
12,470
15,764
19,751
25,008
27,038
0 MW
5000 MW
10000 MW
15000 MW
20000 MW
25000 MW
30000 MW
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Cumulative MW Installed IA Signed-Financial Security Posted IA Signed-No Financial Security
The data presented here is based upon the latest registration data provided to ERCOT by the resource owners and can change without notice. Any capacity changes will be reflected in current and subsequent years' totals. Scheduling delays will also be reflected in the planned projects as that information is received. This chart reflects planned units in the calendar year of submission rather than installations by peak of year shown.
Financial security posted for funding interconnection facilities does not include CREZ security deposits, which are refunded to the Interconnecting Entity when an IA is signed.
ERCOT Wind Installations by Year (as of July 1, 2016)
Wind Growth Potential
PUBLIC 5
Wind Regions within ERCOT
Panhandle:
3,171 MW
South:
976 MW
West:
9,572 MW
North:
1,116 MW
Coast:
2,041 MW
PUBLIC 6
Wind Records
TIMELOAD
(MW)
WIND GENERATION
(MW)
PENETRATION AT
RECORD TIME
02/18/2016 21:20 35,528 14,023 39.47%
03/23/2016 01:10 27,245 13,154 48.28%
PUBLIC 7
Wind Generation Forecast
• AWS Truepower computes a Short Term Wind Power Forecast (STWPF) for all
Wind-powered Generation Resources (WGR) in ERCOT for a rolling 168 hour
timeframe.
• This process uses information from multiple systems, including:
– RARF (Resource Asset Registration Form) (ex. site geo-location, met tower geo-location)
– EMS (Energy Management System) (ex. Resource Status, telemetered site specific
meteorological data)
– Outage Scheduler (ex. WGR derate start/stop dates, derate values)
• When submitting COPs, Wind QSEs must use this STWPF for the applicable
168 hours timeframe. (COP HSL must be <= STWPF.)
– Note that NPRR 785, upon approval will only require Wind QSEs to update a COP only
when wind farm operating conditions dictate COP HSL to be lower STWPF.
*Note that for a WGR going through commissioning process, ERCOT will start providing STWPF to the QSE once WGR receives approval to
Generate into the ERCOT Grid (Part II of the Commissioning check list)
PUBLIC 8
WPF Process Inputs
Source Data
Registration
System
1. Resource Name
2. Location of wind farm (latitude and longitude or equivalent for the center point of wind farm)
3. Location of the meteorological tower (latitude and longitude or equivalent)
4. Type (manufacturer/model) and number of turbines
5. Turbine hub height(s) above ground level with associated number of turbines
6. Manufacturer’s power curve (capability curve)
7. Resource Commercial Operation Date
Energy
Management
System (EMS)
1. Most recent Resource (wind farm) status with date/time
2. Most recent MW output of wind farm with date/time
3. Most recent wind speed and direction at hub height from one meteorological tower with date/time
4. Temperature and barometric pressure at 2 m above ground level on the meteorological tower
Telemetry Value
(Sent every 5
minutes from
EMS)
1. MW Average
2. Wind Speed (mph)
3. Wind Direction (degrees)
4. Temperature (ºC)
5. Barometric Pressure (mbar)
6. HSL Average
7. Num of Turbines ON
8. Num of Turbines Off
9. Num of Turbines Unknown
10. Curtailment Flag
Outage
Scheduler
1. Turbine Outage/Unit Derate Value (HSL)
2. Turbine Outage/Unit Derate Start
3. Turbine Outage/Unit Derate Stop
PUBLIC 9
NPRR 785
• Synchronizes Short Term Wind Forecast (STWPF) with Current Operating Plan
(COP) of Wind Generation Resources (WGRs) and Photovoltaic Generation
Resources (PVGRs)
– ERCOT will automatically prepopulate every WGR and PVGR’s COP with most recent
forecast for the next 168 hours.
– QSEs representing WGRs or PVGRs will be required to either submit a COP with the
automatically prepopulated forecast or submit a lower number (when wind farm operating
conditions dictate COP HSL to be lower STWPF).
PUBLIC 1
0
• Relatively concentrated geographic region for wind installation (lack of spatial
diversity)
• Texas weather can change rapidly
• Severe changes in wind generation output caused by different types of extreme
weather (large wind ramps)
– Frontal system, trough, or dry line/Thunderstorms/Low-level jets/Weakening pressure
gradients/Strengthening pressure gradients
• There is an approximation of hypothetical wind speed-power curve to the real
one
Hypothetical
Power Curve Actual Wind Farm
Power
*The steep part of the power curve from 4-12 m/s is where power
increases strongly with speed.
Challenges in Forecasting Wind
PUBLIC 1
1
Wind Forecast Errors
0
1000
2000
3000
4000
5000
6000
7000
8000
0%
1%
2%
3%
4%
5%
6%
7%
8%
Jul-15 Aug-15 Sep-15 Oct-15 Nov-15 Dec-15 Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16
Mo
nth
ly M
ean
Es
tima
ted
Un
cu
rtaile
d P
ow
er O
utp
ut [M
W]
Mo
nth
ly M
ean
Ab
so
lute
Pe
rce
nt
Err
or
of
Insta
lle
d C
ap
acit
y
WIND POWER FORECAST SYSTEM-WIDE Mean Absolute Percent Error
Monthly Mean Estimated Uncurtailed Power Output [MW] Day-Ahead 1430 STWPF
Day-Ahead 1430 COP HSL Hour-Ahead STWPF
Hour-Ahead COP HSL
––
PUBLIC 1
2
Wind Forecast Errors (Day-Ahead)
––
2013
2014
2015
6.0%5.7%
6.7%6.0%
6.9%
5.2%5.6% 5.8%
5.1%
4.0%
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MA
PE
DAY AHEAD WIND FORECAST PERFORMANCE
2013 2014 2015 2016
MAPE = MEAN ABSOLUTE %ERROR
PUBLIC 1
3
Wind Forecast Errors (Hour-Ahead)
––
2013
2014
2015
3.84%3.52%
3.81%
4.26%4.51%
3.88%3.64%
3.98%
3.40%
2.70%
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC
MA
PE
HOUR AHEAD WIND FORECAST PERFORMANCE
2013 2014 2015 2016
MAPE = MEAN ABSOLUTE %ERROR
PUBLIC 1
4
Wind Generation Cold
Weather Preparation
PUBLIC 1
5
Icing Impacts
Source: Wind Turbines in Cold Climates: Icing Impacts and Mitigation Systems, http://www.springer.com/us/book/9783319051901
PUBLIC 1
6
Wind Units Failure in Cold Winter
• Frigid Temperatures Exceeds Turbine Limits
‒ Wind turbines are typically designed to operate within ambient air temperatures of -15°C/-
20°C (5F/-4F)
‒ Wind turbines have an automatic shutdown feature to protect components if that range is
exceeded.
• Blade Icing
‒ Icing on wind turbines affects three different aspects simultaneously
• the design (aerodynamics, load, control system, and material)
• the safety (ice throw, unbalanced rotor spinning, over-power, and fatigue)
• performance (annual energy output, wind measurements, and design life duration)
‒ Other impacts
• wind sensors—rendering ineffective wind-measuring equipment
• increase noise levels and generally decrease a turbine’s cost-effectiveness.
PUBLIC 1
7
Blade Icing
PUBLIC 1
8
Safety
PUBLIC 1
9
Inclement Weather Plan
Preventative Maintenance
Schedule
Emergency Plan
Communication
Cold Weather Preparation
PUBLIC 2
0
Cold Weather Extreme Package• This ensures that wind turbine operates in temperatures as low as -30C (-22F),
and in survival mode without operation, at temperatures as low as -40C (-40F)
(one example).
• ERCOT is aware of at least one vendor that offers an extreme cold weather
package. Resource Entities should check with their manufacturer.
Minimum temperature (standard)
operational / survival
Standard weather package -15°C/-20°C (5F/-4F)
Cold weather package -30 °C/-40 °C (-22F/-40F)
PUBLIC 2
1
Cold Weather Packages
• “Cold Weather Packages” extend temperature ranges by heating components
such as the nacelle space, yaw drive and pitch motors, and the gearbox, slip
ring, controller and control cabinet, and battery.
• Heating could also prevent another insidious cause of turbine failure.
– When a turbine is not running, oil that is stationary in radiator passages can quickly cool,
and its viscosity can increase.
– Even if wind turbines are not being used, an important lesson worth learning is that the
turbines should be cycled online to provide flow of cooling oil.
– All cooling equipment for radiators on wind turbines should also be disabled for cold weather
events.
PUBLIC 2
2
Icing Prevention Technologies• Passive icing prevention methods
– Rely on the physical properties of the blade surfaces to prevent ice accumulation.
– An example is application of an anti-adhesive coating on the blade such as teflon.
– Another approach takes advantage of the heat absorbing capacity of dark colored surfaces
and consists in the use of black coated blades.
• Active de-icing methods
– Consist of thermal, chemical and impulse de-icing
– In thermal de-icing, electrical elements, similar to the one found on the rear window of a car,
can be used to warm and melt the ice accumulation off the blades.
• In a comprehensive wind turbine icing prevention approach, sensors that could
detect the build-up of ice on the rotor could be considered.
PUBLIC 2
3
Wind Operations During Icing
Conditions
PUBLIC 2
4
Challenges Created by Icing
• Icing creates a mismatch between the forecast and actual production.
• This creates an error in the forecast which then was provided to WGRs.
• This error may be used in the COP which feeds into RUC.
– Could lead to ERCOT counting on more wind than WGRs are able to generate.
PUBLIC 2
5
Icing Event Dec 26-27, 2015• Freezing temperatures on December 26, 2015 in the panhandle region of Texas
marked the beginning of icing conditions that lasted throughout the duration of
December 27.
• By 3:00 AM on December 27, the wind forecasting error had reached about
1700 MW. Icing was continually reported throughout the day and forecasting
error reached nearly 4500 MW by the hour ending HE0800.
• The outage scheduler reported new WGRs related derates and forced outages
totaling around 6000 MW by the end of the day.
PUBLIC 2
6
Wind Outages 12/26 – 12/27
PUBLIC 2
7
West Wind Forecasts Over Time – HE8
Observed: 2,285
Last COP: 5,864
6 hour: 6,937
PUBLIC 2
8
West Wind Forecasts Over Time – HE19
00
Observed: 1,352
Last COP: 1,774
6 hour: 3,110
12 hour: 5,592
PUBLIC 2
9
West Wind Hour Ahead Summary
0.0
1,000.0
2,000.0
3,000.0
4,000.0
5,000.0
6,000.0
7,000.0
8,000.0
9,000.0
10,000.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Forecast Error STWPF Aggr COP RT Aggr Wind-Output
PUBLIC 3
0
Panhandle Wind Hour Ahead Summary
0.0
500.0
1,000.0
1,500.0
2,000.0
2,500.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Forecast Error STWPF Aggr COP RT Aggr Wind-Output
PUBLIC 3
1
Forecast vs. COP Comparison
0.0
1,000.0
2,000.0
3,000.0
4,000.0
5,000.0
6,000.0
7,000.0
8,000.0
9,000.0
10,000.0
1 3 5 7 9 11 13 15 17 19 21 23
West Wind Hour Ahead Summary
Forecast Error
STWPF Aggr COP
RT Aggr Wind-Output
0.0
500.0
1,000.0
1,500.0
2,000.0
2,500.0
1 3 5 7 9 11 13 15 17 19 21 23
Panhandle Wind Hour Ahead Summary
Forecast Error
STWPF Aggr COP
RT Aggr Wind-Output
PUBLIC 3
2
Notice the disparity between the last forecast value before the adjustment period and the observed value in HE 8 and 12
versus in HE 19. Delayed outages caused these errors in earlier hours which were then minimized in HE 19. The faster
these outages are entered, the faster they are incorporated to the forecast.
Forecast Performance
0
500
1000
1500
2000
2500
HE 3 HE 8 HE 12 HE 19
5 Hour Forecast MW
Last Forecast Before Adjustment Period MW
Observed
Panhandle Wind West Wind
0
1000
2000
3000
4000
5000
6000
7000
8000
HE 3 HE 8 HE 12 HE 19
5 Hour Forecast MW
Last Forecast Before Adjustment Period MW
Observed
PUBLIC 3
3
Unit’s STWF with no Outage Updates
Forecast NEVER
Converged to zero
WRONG
PUBLIC 3
4
Unit’s STWF with Outage Updates
Forecast Converged
to zero
CORRECT
PUBLIC 3
5
Telemetry During Icing Conditions
Number of
online turbines
Number of
offline turbines
MW output
*Remember this is under icing
conditions.
While the NTON line does
decrease, it does not fall as the
unit’s MW goes to zero, but is
very delayed.
This is too slow.
WRONG
PUBLIC 3
6
Telemetry During Icing Conditions
Number of
online turbines
Number of
offline turbines
MW output
*Remember this is under icing
conditions.
The NTON line falls as the unit’s
MW goes to zero.
This is very good.
CORRECT
PUBLIC 3
7
Addressing the Wind Turbine Icing
• Use the following when a WGR is experiencing an icing related event,
– Call the ERCOT Control Room and notify them.
– Continuously update the telemetry for number of turbines on and off (NTON and NTOFF) as
conditions on the ground deteriorate.
– If the outage/derate is expected to last greater than 2 hours, submit this into ERCOT’s
Outage Scheduler.
• In comparison to turbine telemetry, Outage submissions have a greater impact in
correcting the forecast.
3.1.4.5 Notice of Forced Outage or Unavoidable Extension of Planned or Maintenance Outage Due
to Unforeseen Events
(2) Any Forced Outage that occurs in Real-Time must be entered into the Outage Scheduler if it is to
remain an Outage for longer than two hours.
PUBLIC 3
8
RRGRR-0003• A survey was conducted in 2015 to gather additional information to gauge the
operational limits of the various WGRs in ERCOT.
‒ Maximum Operating Temperature (Fahrenheit (°F))
‒ Minimum Operating Temperature(Fahrenheit (°F))
‒ High Wind Speed Cut-Out (meters per seconds (m/s))
‒ High Wind Speed Cut-Out time (minutes)
‒ High Wind Speed Cut-Out Reset (meters per seconds (m/s))
‒ High Wind Speed Cut-Out Reset Time (minutes)
• This information is now part of the RARF and is fed into the wind forecast
received from AWS.
PUBLIC 3
9
Current and Ongoing
Telemetry Concerns
PUBLIC 4
0
Meteorological Telemetery
Telemetry Name Data Unit
Wind Speed (MPH) MPH
Wind Direction (DEG) Degrees
Temperature (TEMP) ºC
Barometric Pressure (PRES) mbar
Num. Of Turbines ON (NTON)
Num. Of Turbines OFF (NTOF)
Num. Of Turbines Unknown (NTUN)
Integer (Count of turbines)
PUBLIC 4
1
Telemetry Examples (from 9/6/2016)
41
PUBLIC 4
2
0
20
40
60
80
100
120
140
0
5
10
15
20
25
30
35
40
06
:00
06
:17
06
:35
06
:53
07
:10
07
:28
07
:46
08
:03
08
:21
08
:39
08
:56
09
:14
09
:32
09
:49
10
:07
10
:25
10
:42
11
:00
11
:18
11
:35
11
:53
12
:11
12
:28
12
:46
13
:04
13
:21
13
:39
13
:57
14
:14
14
:32
14
:50
15
:07
15
:25
15
:43
MW HSL NTON NTOF
• Wind farms interpret the NTON flag differently.
• The NTON is meant to describe the number of turbines available for production,
and has nothing to do with the actual output.
– The left image is taking the unit’s MW value into account - Wrong
– The right image is expected method for computing and telemetering this data.
Number of Turbines Online (NTON)
0
5
10
15
20
25
30
35
06
:00
06
:16
06
:33
06
:50
07
:06
07
:23
07
:40
07
:57
08
:13
08
:30
08
:47
09
:04
09
:20
09
:37
09
:54
10
:11
10
:27
10
:44
11
:01
11
:17
11
:34
11
:51
12
:08
12
:24
12
:41
12
:58
13
:15
13
:31
13
:48
14
:05
14
:22
14
:38
14
:55
15
:12
15
:28
15
:45
MW HSL NTON NTOF
CORRECTWRONG
0
5
10
15
20
25
30
350
6:0
00
6:1
60
6:3
30
6:5
00
7:0
60
7:2
30
7:4
00
7:5
70
8:1
30
8:3
00
8:4
70
9:0
40
9:2
00
9:3
70
9:5
41
0:1
11
0:2
71
0:4
41
1:0
11
1:1
71
1:3
41
1:5
11
2:0
81
2:2
41
2:4
11
2:5
81
3:1
51
3:3
11
3:4
81
4:0
51
4:2
21
4:3
81
4:5
51
5:1
21
5:2
81
5:4
5
MW HSL NTON NTOF
WRONG
0
20
40
60
80
100
120
140
0
5
10
15
20
25
30
35
40
06
:00
06
:17
06
:35
06
:53
07
:10
07
:28
07
:46
08
:03
08
:21
08
:39
08
:56
09
:14
09
:32
09
:49
10
:07
10
:25
10
:42
11
:00
11
:18
11
:35
11
:53
12
:11
12
:28
12
:46
13
:04
13
:21
13
:39
13
:57
14
:14
14
:32
14
:50
15
:07
15
:25
15
:43
MW HSL NTON NTOF
The NTON line stays constant despite
the unit’s MW value going to zero.
CORRECT
PUBLIC 4
6
0
20
40
60
80
100
120
140
00
:00
00
:37
01
:15
01:5
302
:31
03:0
90
3:4
70
4:2
505
:03
05:4
106
:19
06
:57
07
:35
08:1
308
:51
09:2
91
0:0
6
10
:44
11:2
212
:00
12:3
81
3:1
6
13
:54
14:3
215
:10
15:4
81
6:2
6
17
:04
17:4
218
:20
18:5
81
9:3
5
20
:13
20
:51
21:2
922
:07
22
:45
23
:23
MW HSL
MW exceeding HSL telemetry• MW telemetry must never exceed the HSL.
• When a WGR is UNCURTAILED its MW and HSL values should be exactly the
same.
WRONG
PUBLIC 4
7
0
20
40
60
80
100
120
00
:00
00
:41
01
:22
02
:03
02
:44
03
:26
04
:07
04
:48
05
:29
06
:10
06
:52
07
:33
08
:14
08
:55
09
:36
10
:18
10
:59
11
:40
12
:21
13
:02
13
:44
14
:25
15
:06
15
:47
16
:28
17
:10
17
:51
18
:32
19
:13
19
:54
20
:36
21
:17
21
:58
22
:39
23
:20
MW HSL
MW exceeding HSL telemetry• MW telemetry must never exceed the HSL.
• When a WGR is UNCURTAILED its MW and HSL values should be exactly
the same.
CORRECT
PUBLIC 4
8
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0
10
20
30
40
50
60
70
80
19:0
0
19
:03
19
:07
19:1
1
19:1
5
19
:19
19:2
2
19:2
6
19:3
0
19
:34
19:3
8
19:4
1
19
:45
19
:49
19:5
3
19:5
7
20
:00
20
:04
20:0
8
20:1
2
20
:16
20:1
9
20:2
3
20:2
7
20
:31
20:3
5
20:3
8
20:4
2
20
:46
20:5
0
20:5
4
20
:57
MW HSL Curtailment Flag
WGR Under Curtailment
• ERCOT nodal protocols and Business Practices dictate that when under
curtailment a WGR’s HSL must represent actual power generation potential.
WRONG
• HSL should never
lower dramatically
under curtailment
scenarios.
PUBLIC 4
9
WGR Under Curtailment• ERCOT nodal protocols and Business Practices dictate that when under
curtailment a WGR’s HSL must represent actual power generation potential.
CORRECT
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0
20
40
60
80
100
120
140
160
12
:00
12
:28
12
:56
13
:24
13
:53
14
:21
14
:49
15
:17
15
:46
16
:14
16
:42
17
:10
17
:39
18
:07
18
:35
19
:04
19
:32
20
:00
20
:28
20
:57
21
:25
21
:53
22
:21
22
:50
23
:18
23
:46
00
:14
00
:43
01
:11
01
:39
02
:08
02
:36
03
:04
03
:32
MW HSL Curtailment Flag
• HSL should
never lower
dramatically
under
curtailment
scenarios.
PUBLIC 5
0
WGR Under Curtailment• WGRs must follow ERCOT Base Point when under curtailment
CORRECT
100
105
110
115
120
125
130
135
140
145
150
18:0
0
18:0
3
18:0
7
18:1
1
18:1
5
18:1
9
18:2
2
18:2
6
18:3
0
18:3
4
18:3
8
18:4
1
18:4
5
18:4
9
18:5
3
18:5
7
19:0
0
19:0
4
19:0
8
19:1
2
19:1
6
19:1
9
19:2
3
19:2
7
19:3
1
19:3
5
19:3
8
19:4
2
19
:46
19
:50
19
:54
19
:57
MW HSL Base Point
PUBLIC 5
1
Questions?
Top Related