20060308Comments.pdf

cd/A'-" ~~JOlt
3jf61
Jean Jewell
110 A.lJ /10
1-1
From:
Sent:
To:
Subject:
Ed Howell
Wednesday, March 08 , 2006 3:53 PM
Jean Jewell; Ed Howell; Gene Fadness; Tonya Clark
Comment acknowledgement
WNW Form Submission:
Wednesday, March 08, 2006
3: 52: 58
Case: pac-E-05-9
Name: Gary Seifert
Street Address: PO 1625, MS 3810
City: Idaho FallsState: id
ZIP: 83415-3810
Home Telephone: [phone]E-Mail: sei~inel. gov
Company: PacifiCorp
mailing list yes no: yesComment=descriptIon: Comments on Schwendiman;
I am an electrical power engineer at the INL.
and 25+ years of working in this field.My comments are based on my knowledge base
I have reviewed this contract. You have my comments on file from the earlier review of
the contract you rejected. None of these have changed and all still apply. Following are
my comments for this submittal, Case PAC-E-05-9.1. The inclusion of the 90/110 banding meets your requirements of rejection. This
language will impact cash flow for Schwendiman and in effect reduce the cost of
electricity for ratepayer s, a point I have to assume you were trying to achieve with yourearlier rej ection.2. Because this contract meets your requirements, I strongly suggest approving it.3. That said, I have two more comments that are aimed at long range wind contracts inIdaho. First is the mis-perception that 90/110 will ensure better forecasting and be an
incentive for performing and delivering power. This is incorrect. All it does is create anegati ve incentive for forecasting. Because of the penal ties, this contract will always
receive low performance estimates. Based on discussions with over half of the people
looking at Idaho PURPA's, they all plan on projecting power with at least a 10% margin to
avoid hitting the 90% band number and the onerous penal ties incurred. (ie - forecast
their power at 90% as 100% to give a margin). There have been rate cases and analysis
performed in regional states, and all have one common conclusion. Anytime forecasts are
too low , the utility places more power on reserve and pays for that cost of spinningreserve and as a rule never uses that power. IE - spending money for no added value.
Therefore, the 90/110 band penalizes the wind farmer and the utility. The utility has to
recover costs and overhead and therefore, the cost eventually get's blended in and passed
on to the consumer. Ergo - the 90/110 band costs both ratepayers and wind farmers, not
what I assumed that you planned with this incentive based rate. Therefore, using a system
like the 90/110 band means that the utility is always paying too much if they follow the
forecasts. And if they do not follow the forecasts then what's the value of the band and
why pay to administer a band that brings no value to Idaho ratepayers?4. The real issue is good forecasting. All utili ties use accurate and real time
forecasting methods where monthly, weekly, daily, and hour ahead numbers are evaluated and
wise decisions made based on real numbers constantly adjusted due to daily conditions.
Both Pacificorp and Idaho Power have modern powerful SCADA systems. Both SCADA systems
have the option to integrate forecasting tools. Current wind forecasting is friendly withutili ties needs for weekly, daily, hourly forecasting and offers accuracies well wi thin
normal estimating limits currently used. Wind forecasting at the utility level exceeds
the level of confidence they have on forecasting consumer load and water flows. If they
are uncertain about integrating these tools, then commercial services are available.Regardless, I assume that Idaho utili ties would rather have control of their forecasting
tools rather than depending on a small agriculture based entity.
Comment summary -
kill 90/110 and direct the utilities to perform their own forecasting and
put a maintenance-based incentive on the wind farm to keep them operational.old MAG approach appears to be a better deal for the ratepayer.The
Thank you for your time and consideration.
Gary Seifert EE PE
INL, PO Box 1625 - MS 3810
2525 Fremont Ave.
Idaho Falls, Id. 83415-3810Email gary. seifert~inl. gov
Phone [phone]Cell 208-521-8385
Fax [phone]
Transaction ID: 381552.
Referred by: http: / /www.puc. state. id. us/ scripts/polyform. dll/ ipuc
User Address: 134.20.
User Hostname: 134.20.
~/O"
Nerg"
~~
VISION LLC
v1iJ AI!/1o
Glenn S. Ikemoto
Principal
Envisioning sustainable future
672 Blair Avenue
Piedmont, CA, USA 94611
[phone] :T
[phone] :F
glenni(g)pacbell.net :E
VIA EMAIL
March 8 , 2006
President Paul Kjellander
Commissioner Dennis Hansen
Commissioner Marsha Smith
Idaho Public Utilities Commission
472 West Washington Street
PO Box 83720
Boise, 1083720-0074
Re: Comments on PAC-05-
Dear Commissioners:
In late December, the developers of the Schwendiman Wind Project asked me if I would
help them negotiate a compromise 90/110 mechanism, as directed by the Commission.
That effort resulted in the contract you are now being asked to approve. I believe the new
mechanism is far more economically efficient than the prior version. That is, delivering a
dollar of savings to the ratepayer will cost the project a dollar. Because of price volatility,
there is a much larger cost to the projects than associated benefit to the ratepayers under
the prior mechanism. In the long run, this loss is not in either party s interest.
My real purpose in writing this letter is to commend the efforts of PacifiCorp
representatives and the Commission s Staff. I have negotiated power contracts from both
the utility and developer sides of the table for over 25 years. This negotiation was one of
the most positive and reasonable I have ever been involved with. I would like to
particularly highlight the efforts of Bruce Griswold of PacifiCorp. He was extremelyresponsive and gave up many hours on nights and weekends to meet a very tight
schedule. He conducted himself in good faith and effectively focused the resources of his
company on solving problems as soon as they arose.
Your Staff also provided critical support at key times during the negotiation. While they
made it clear that they would not dictate an outcome or prejudge any results, they
unfailingly urged the parties to seek compromise. Without this moral support, I don
believe we would have succeeded. You can be justifiably proud of their professionalism.
The Schwendiman project deserves to go forward. The development team has invested
years of effort bringing the project to this point. This time and money was invested in thereasonable belief that a market for clean energy existed in Idaho. For the developers, the
final form of the contract does not really represent a compromise, it is a pure concession.
The new 90/110 mechanism will result in an unavoidable reduction in project revenue
compared to the original contract. There was only a small overlap between the
requirements of the project's financiers and the Commission s directions. I hope we have
succeeded in finding it.
st re ards '
Glenn Ikemoto
Principal
'3J~Jov
Jean Jewell
~~
All l1'tI
;.:
From:
Sent:
To:
Subject:
Ed Howell
Wednesday, March 08 , 2006 12:54 PM
Jean Jewell; Ed Howell; Gene Fadness; Tonya Clark
Comment acknowledgement
WWW Form Submission:
Wednesday, March 08, 2006
12:53:51 PM
Case: PAC-E-05-09
Name: Paul Dawson
Street Address: 5180 S. Forest Floor Ave.
City: Boise
State: ID
ZIP: 83716
Home Telephone: [phone]
E-Mail: pdawson~boisestate. edu
Company: Idaho Power
mailing list yes no:
Comment=descriptIon: ~m a professor of Mechanical Engineering at
University. I have a background in Meteorology and thermal-fluid sc
Engineering. I wanted to offer a comment on the 90/110 rule.
Boise State
iences in Mechanical
As a research professor who has studied meteorological modeling, I know how difficult it
is to forecast weather, particularly winds in the atmospheric boundary layer. I have been
trying to forecast temperature, relative humidity, and winds in the Treasure Valley of
Idaho during winter for the last 3 winter seasons. The above-surface winds tend to be the
most difficult to forecast, even one day in advance, using the state-of-the-science
mesoscale weather research and forecasting (WRF) model, currently being used in the
research and forecasting communi ties. Trying to provide an accurate wind forecast is
difficult, if not impossible, beyond 48 to 72 hours. The forecast gets more difficult the
longer the forecast period. For a forecast 3 months in advance, the best approach is to
use statistical data and statistical averages. But this data varies from year to year and
is influenced by phenomena such as El Nino and climate change dynamics. The latter
phenomena tend to make statistical studies less reliable, as recent studies and journalpublications indicate.
I hope this comment has been helpful in judging whether the 90/110 rule is a fair rule forwind energy proponents and developers.
Paul Dawson,PhD
Transaction ID: 381253.
Referred by: http: / /www.puc. state. id. us/ scripts/polyform. dll/ipuc
User Address: 132.178.127.
User Hostname: 132.178.127.
I~IO
Jean Jewell
((0 AY'
$ ~ ttIJ;
From:
Sent:
To:
Subject:
Ed Howell
Wednesday, March 08, 2006 9:07 AM
Jean Jewell; Ed Howell; Gene Fadness; Tonya Clark
Comment acknowledgement
WNW Form Submission:
Wednesday, March 08, 2006
9:06:30 AM
Case: PAC-E-05-09
Name: Dr. David Lubitz
Street Address: 85 Behr Ave #204
City: San Francisco
State: CA
ZIP: 94131
Home Telephone: [phone]
E-Mail: wdlubitz~ucdavis. edu
Company: Idaho power
mailing list yes no: yes
Comment descriptIon: am submitting the comments below in regard to case PAC-E-05-09. Iam commenting specifically on the 90/110 banding requirement included in the agreement. Isubmit these comments as an individual with an extensive background in wind energy
forecasting, and an interest in encouraging economical energy production in the state of
Idaho.
I have had extensive experience forecasting the energy production of wind plants during
the last several years. There has been a long term interest in wind energy forecasting in
California, and the California Energy Commission (CEC), in partnership with the ElectricPower Research Institute (EPRI) have supported two multi-year projects to improve theforecasting of wind energy production at wind plants in California. These studies
contracted with leading wind energy forecasting companies and the University ofCalifornia, Davis, to produce energy forecasts for several wind plants in different
regions of California. I performed research at the University of California, Davis insupport of these studies, and also designed an operational wind energy forecasting system.The goal of these projects was to improve the accuracy of "short-term" forecasting, whichis the prediction of wind energy production between zero and 48 hours in advance. Theseprojects have documented a marked improvement in wind energy forecasting accuracy, fortime periods up to 48 hours in advance. The results of the first proj ect have beenpublished by EPRI (EPRI Technical Report 1007339. California Wind Energy Forecasting
System Development and Testing. Phase 2: 12-Month Testing. Palo Alto, CA, USA, July 2003)and the final report for the more recent study, the California Regional Wind EnergyForecasting System Development proj ect, is in preparation and will be published shortly.These studies have documented that wind energy forecasting up to 48 hours in advance hasimproved markedly in California. Other research has suggested that the improvements seenin California short-term forecasting are typical of the improvements in the short-term
forecasting ability of the forecasting industry as a whole.
It should be noted that the wind energy forecasting industry is focused on producing
forecasts between zero and 48 hours in advance, typically for hourly power production.
Most of the demand from the power generation industry has been for forecasts in this timeframe, as the "next hour" and "next day" forecasts provide a good combination of forecast
accuracy and operational usefulness. The next day forecasts are used to assist in
contracting of power and planning generation , while forecasts between zero and three hours
in advance are used by power grid operators to balance generation and demand. Forecastingmore than 48 hours in advance is difficult to do with useful accuracy, and has lesspractical value for most stakeholders. Because of this, there is little research ordevelopment focused on forecasting wind energy production weeks or months in advance, andfew forecasting products are available for this type of forecast.
I believe the 90/110 banding requirement places a significant burden on wind plant
operators. Based on my experience, forecasting the monthly energy production of a wind
plant a quarter in advance, while consistently guaranteeing forecast error of less than
10% on a monthly basis, is not possible with current forecasting tools. Given thedifficulty of weather forecasting in general over these time frames, it is unlikely that
this will be achievable in the near term. Since the wind energy forecasting industry has a
variety of products available for shorter-term forecasts that are continually improving in
accuracy, consideration should be given to replacing 90/110 banding with a requirement for
next day energy forecasting. The next day forecasting products available may provide a
better combination of greater operational benefit, accuracy and cost-effectiveness, giventhe accuracy limitations of longer-term forecasts and the difficulty of producing them. Ifthe 90/110 banding is retained, it might be possible to reduce the burden on wind plant
operators by making further adjustments to the 90/110 requirements, such as allowing alarger error band or changing the financial incentives.
Thank you.
David Lubitz, PhD
Transaction ID: 38906.
Referred by: http: / /www.puc. state. id. us/ scripts/polyform. dll/ipucUser Address: 71.139.185.189
User Hostname: 71.139.185.189