Monday, July 6, 2009

Electricity in NH - Wind Power

Lempster Wind project in Lempster, NH (photo courtesy of wikipedia)

David Brooks, of and the Nashua Telegraph, recently published a post on the Vinalhaven, ME,  Fox Islands Wind project.  His post got me digging into wind power a bit and I figured now was as good a time as any to do my first post on it.

I did some quick calculations on the Fox Islands Wind project and I realized right away that this project was much more costly than most wind power projects.  Wind projects usually cost around $2,000 per kilowatt to build, but the Fox Islands project is projected to cost over $3,200 per kW.   Because of its remote location, the Fox Islands project is priced more like an offshore wind project than an onshore project.  Offshore projects are more expensive because of the extra cost of transporting equipment and workers from the mainland and due to costs of erecting wind towers on the water.  

Normally, a wind project this expensive would be a tough sell, but because of Vinalhaven's remote location, residents pay a lot for electricity from the mainland.   Most folks in New England pay around 13-15 cents per kilowatt hour for electricity.  Vinalhaven residents pay 23 cents.  That allows the Fox Islands Wind project to be competitive, despite its higher cost.

It turns out that the midwestern states have the most onshore wind potential of anywhere in the US.  Offshore, there are strips of great potential along the east and west coasts and in the great lakes region.   Currently, Texas leads the nation in wind power generation, with nearly 8,000 MW of generation capacity installed. 

For its part, New Hampshire isn't particularly well suited for wind power overall.  This is despite the fact that the highest wind speed ever recorded on the surface of the earth was measured at the top of Mount Washington.  Although New Hampshire doesn't have vast amounts of onshore wind potential, we do have a few concentrations of high potential, particularly in the mountains.

With only small pockets of high potential, the trick for New Hampshire and New England is to identify the sites that make the most sense to develop.  To see how wind power developers go about doing this, we'll explore the finances of a few active projects.  The best-practice approach for analyzing the finances of large capital projects, including wind power projects, is to use discounted cash flow (DCF) analysis.   This methodology takes into account the fact that a dollar received in 1 year is worth more than a dollar received in 20 years.  

DCF models can be hand-built in MS Excel, but fortunately I found a great online tool called WindFinance that was created by the National Renewable Energy Laboratory (NREL).  NREL is the U.S. Department of Energy's primary research lab for renewable energy and energy efficiency.  The WindFinance tool is a very sophisticated financial modeling tool and it was relatively easy to use and understand.

I used the tool to create models for a few local wind projects.  Since I don't have detailed cost information for these projects, I used estimates from publicly available sources and historical cost data from NREL.  Because some of these inputs are just guesses, the tool's outputs are probably not very precise. That's ok, since the real goal of the exercise is to show the basic economics of wind power and show some of the factors that can impact the cost of energy for a project.

The projects I chose to evaluate are the Lempster Wind project, in Lempster, NH which began operation in 2008, the Coos County Wind project, which is still in the planning phase, and the Fox Islands Wind project, in Vinalhaven, ME, which was mentioned in David Brooks' post.

The table above shows the values I used as inputs in the analysis.  These inputs should be pretty self explanatory, especially if you've read my earlier post about electricity generation economics.   Installed cost per kW and annual production in MWh are actually computed from the other inputs.  The installed cost per kWh for the Coos County project is on the high side compared to Lempster, and compared to national averages.  This is probably  because the turbines will be installed along mountain ridges at high elevation.  Coos is priced about halfway between the average cost of onshore and offshore wind projects.

The "Operating cost per kW installed" input was basically a guess.  This input is the total annual cost for fixed and variable expenses associated with running the wind farm.  It includes things like maintenance on the turbines, property taxes, insurance, and any other operating overhead.  For simplicity, I combined all these costs and modeled them as an expense based on kilowatts of installed capacity.  Also, I used a life expectancy of 20 years for all scenarios.

The last column of the table above shows the WindFinance tool's estimate for the levelized cost of energy for each scenario that I modeled.  The LCOE output represents the minimum wholesale price that a power generator can receive for their output and still cover all their expenses.   I ran a few scenarios with variations in cost of capital, tax benefits, and renewable tax credits, in order to show how changes in these inputs can impact the cost of energy. 

The dollars per MWh and cents per kWh estimates in the table above represent levelized wholesale energy costs.  To put these costs into perspective, exhibit G-13 of PSNH's LCIRP, estimated 2009 energy prices at between $47 and $63 per MWh for off peak power and between $64 and $85 per MWh for on peak.  

For the baseline cases above, I assumed a $21 per MWh renewable energy production tax credit (PTC) for 10 years.   Recent federal legislation allows developers to choose between the production tax credit or an investment tax credit of 30% of the capital cost of the project.     Incidentally,  the subsidies for wind are much greater in Europe.  In Germany for example, the credit is currently set at $90 per MWh for the first 5 years, then $50 per MWh  thereafter.

Also in the table above, you can see how sensitive the final energy cost is to changes in cost of capital, tax handling, and tax credits.  For the Fox Islands project, I modeled the project structured as a corporate for-profit project, probably using equity financing, and as a community or utility project, perhaps financed mostly with debt.  For the first case, I used a 13% cost of capital with a 40% tax rate and for the second I used an 8% cost of capital and a 0% tax rate.  This simple example really shows how the structure for the project, eg  independent power producer vs. utility, vs. community project can really impact the economics.

With this analysis, you can see that wind power isn't the panacea of plentiful, dirt-cheap electricity that some make it out to be.  OTOH, when the costs of power plant emissions and our fossil fuel dependence are brought into the equation, wind power has proven itself to be a viable part of the solution.  In addition, technological advances continue to improve wind power economics and the pace of innovation shows no signs of slowing.

My analysis didn't take transmission costs, capacity payments, or firming costs into account.  Firming costs are often assessed on wind generators because their output can be unsteady and may have negative impacts on the power grid.  All three of these factors can have an impact on the final delivered cost of energy from wind power.

To wrap up, below is one last table that shows basic cost and estimated production information for some active wind projects in New Hampshire and Maine.  Maine currently has the most wind power generation in New England and has been actively encouraging wind power projects.  The state recently passed legislation to fast-track permitting for wind projects located inside certain remote areas of the state that officials felt are well-suited for utility-scale wind projects.

(Note: I corrected a mistake I made in estimating the capacity factor for the Lempster Wind project.  All the tables now reflect the correct value.)



  1. At the risk of blog-rolling, great analysis. As soon as I can, I will steal it - er, discuss/link to it!

    Dave Brooks

  2. By the way, this post is now the official reference in the Wikipedia article "Wind Power in New Hampshire" for the Lempster Mountain annual output figure. So you have to maintain this blog forever, or disappointed Wikipedia hordes will descend upon you!

  3. Dave,

    Thanks for the comments and the link-back on Wikipedia.

    Once I saw you were quoting the Lempster annual production, I went back to re-check where I got that. I thought for sure I read it was 30% somewhere, but to be sure I went back to the PUC filings and saw that Lempster Wind LLC was projecting 33%. That raises the annual production from 63k to 69k.

    Anyhow, I updated all my figures above to reflect the correct 33% capacity factor and I'll update the wikipedia article as well. Sorry about that!


  4. Based on an email I got from the folks at Iberadola Renewables (the company that build the Lempster Wind project), I've updated the Lempster Wind capacity factor again. The 33% number I used was from some PUC filings by PSNH, but Iberdrola tells me that 36-37% is the correct number.

    I'll also fix the wikipedia article as well.


  5. Kibby Mountain Wind Turbine

    The 132 MW wind turbine facility, capital cost about $330 million, is
    owned by TransCanada and was built, after a lot of destruction, on one
    of the most beautiful ridge lines in Maine. TransCanada and Vestas the
    claimed that the capacity factor would be 0.32, or greater. It was
    placed in service on 10/31/2009.

    Its FERC designation is Trans Canadian Wind Development, Inc. in case
    you want to look up the below data.
    Here is the FERC site

    In 2009 and 2010, the facility had a lot of startup problems and its
    energy production was negligible.
    In 2011, it had a capacity factor of 22.5% for the first 9 months.
    For the 3rd quarter of 2011, it was 14.42%. Monthly capacity factors
    were as follows:
    July 18.48%
    Aug 12.31%
    Sept 12.41%

    Why are the CFs so low?

    Winds on ridge lines have highly-irregular velocities AND directions.
    This does not show up when one does wind velocity testing for
    feasibility, but when rotors are 373 feet in diameter, one part of a
    rotor will likely see a different wind velocity AND direction from
    another part. This leads to highly inefficient energy production and
    CFs. Wind vendors are very familiar with this, but do not mention it.
    However, all is explained in this article. I recommend the VT-DPS and
    House Environment and Energy Committee, and all others, finally read
    this article, before "leading" Vermont into an expensive energy


Note: Only a member of this blog may post a comment.