This article has been cross-posted on The Oil Drum.
Last month, I brought you some nice
maps showing when and where good wind resources are found in the US.
Now I've found something better: a visual
comparison of electrical load with wind farm production[pdf file], published
by the Western Area Power Administration in 2006. The study compared electricity
production from five wind farms in Northern Colorado, Southwestern Nebraska, and
Central Wyoming in 2004, 2005, and the start of 2006, compared with electricity
consumption in the same area over the same time period.
Comparison of Wind Production to Electricity Demand
I've copied four of the most representative graphs below.
The first and third heat graphs below show electricity production at the five
wind farms studied in 2004 and 2005, respectively. The Second and fourth
show electricity demand in the surrounding territory. Red(blue) denotes
areas of high(low) production or demand.
For wind advocates, these are probably rather scary graphs. The first
thing you probably noticed was the big blue patches of wind production during
summer peak demand, roughly 10am to 10pm in June, July, and August.
This is why wind is referred to as an "energy resource" not a
"capacity resource." Right when demand is highest (namely hot summer
afternoons), the
wind is least likely to be blowing.
On Second Thought - How Much Backup Do You Need?
That is just the first impression, and while it is a true impression, it's
also an oversimplification. If you look at the scale, you will notice that
the blues on the wind production graphs actually represent wind generating at
10% to 15% of nameplate capacity. If you factor in the fact that a normal
capacity factor for wind is about 25-40%, that means that even on these hot
summer afternoons, the farms are generating at one-third to one-half of their
"normal" output. This means that, contrary to popular
misconception, wind does not require a "100% back-up with natural
gas." It is true that wind is less reliable than baseload power
plants such as coal and nuclear, which typically run about 90% of the time, but
in an apples-to-apples comparison, a 100 MW coal or nuclear plant will produce
as much energy over the course of a year as a 270 MW wind farm. During the
peak summer months, the coal plant will need some backup power in case of an
unscheduled shut down due to lack available coal (this
happened in Colorado in 2005 due to problems with dust in rail tracks) or lack
of available cooling water during a heatwave, and when a coal or nuclear
plant goes down, it goes all the way down, so the 100 MW baseload plant has a
small chance of needing 90 MW of backup to produce at its "normal"
rate of power production. On the other hand, the wind farm will be operating
at (a conservative) third of its "normal" capacity, producing about
30MW. To bring that up to it's normal capacity for the year, it will need
60MW of back-up power.
In other words, because some part of a large distributed group of wind farms
is always producing some power, it will never go completely down. A large
baseload power plant, on the other hand, is completely down about 10% of the
time (although less during peak summer months, because utilities schedule maintenance
in off seasons.)
Pick Farms to Match Your Load
Another point worth noting, is that the wind has different annual patterns in
different locations. The smallest (8.4 MW out of 139MW) of the five farms
in the study was "Wind Farm B" in central Wyoming. If you look
at the following two heat maps below for 2004 and 2005, which show the
production of just this wind farm, you will note that during the peak summer
demand, this farm was producing at over 50% of "normal" capacity for
much of the summer peak.
Since we know what electricity demand looks like, if we plan new wind farms
(and adequate
transmission), we can choose to build wind farms that produce more power
when we most need it. If all the farms in the example in the last section
had more favorable production patterns like Farm B, even less back-up generation
would be needed to bring them up to "normal" capacity.
For instance, in the Texas Competitive Renewable Energy Zones study [.pdf
7.64MB] wind in the coastal area (along Texas's southern gulf coast) was found to be a much better match for
the ERCOT load shape than wind in other areas, although the average capacity
factor was considerably lower than panhandle wind. See chart below.
Hence, careful selection of wind farms can lead to wind production with
higher capacity during peak loads, and correspondingly less need for
dispactchable power. Although Texas is currently focusing on developing
wind farms in West Texas and the Panhandle because of their high capacity factors
and correspondingly high annual energy output, the power from coastal wind farms
is likely to become increasingly valuable as wind reaches higher penetration.
It's Not All About Summer Peak
Statements about wind's need for large dispacthable backup generation because
of low capacity factors during peak times contain am implicit assumption that
electricity demand is fixed. This assumption is both false and pernicious,
because shifting demand can be done cheaply, and often produces multiple
benefits. While it is true that most large scale electricity storage
technologies, such as pumped
hydropower, compressed
air energy storage, and utility
scale batteries are expensive or limited to a few available sites (pumped
hydro,) technologies which shift the demand curve are not.
If you look back at the first set of four heat maps, you will note that wind
actually does a quite good job serving the winter peak. In 2004 (a
year with a moderate summer) winter peak demand actually exceeded summer
peak.
Capacity during winter peak has some advantages over summer
peak. First of all, natural gas prices are higher during the winter,
because natural gas is used extensively for home heating as well as power
generation. In February 2006, Xcel
Energy had a series of major power outages in Northern Colorado which they
blamed on insufficient natural gas in storage due to an unusually cold
temperatures. Yet as this heat map 
shows, wind farms in the region were operating at 40-60% capacity factors
(i.e. well above "normal" production) for
January and February. Note that the blue at the end of the year was due to lack
of data, not lack of production. Had there been more wind farms
installed, this would have had a large impact on the amount of natural gas
needed for electrical generation, and the outages would not have
happened. I don't have data to back it up, but my personal
experience leads me to believe that cold winters in the great plains are also
particularly windy winters, meaning that winter wind capacity is ideally suited
to displace natural gas needed for heating.
How Heat Pumps Fit In
Which brings me to the title of this article: why heat pumps are an excellent
fit with wind generation. In my article on how
to invest in the Pickens Plan, I mentioned that ground-source
heat pumps (GHP) can displace gas used for heating with a smaller amount of
electricity from wind. Since a GHP is both an efficient air
conditioner as well as an efficient heat source, it not only reduces natural
gas used for heating, but also reduces electricity used for cooling in hot
summer months, which in turn reduces summer peak loads.
Deployment of GHPs does
three things to make energy supplies fit energy demand:
- Winter electricity usage is increased just when wind capacities are
highest.
- Summer electricity consumption is decreased when wind capacities are
lowest.
- Use of natural gas for heating is reduced during times of peak gas demand.
GHPs, because of their extreme efficiency, also have the benefit of saving
users a lot of money.
The Dual Fuel Option
Unfortunately, GHPs have not been widely adopted, due to the difficulties of
installing the buried heat exchange loops, especially in urban areas (although some
utility programs have been very successful.) When I bought a house, it
was in a New Urbanist development with
very small lots which was close to my work. While this saves me countless
gallons of gasoline, it meant that I was unable to use a heat pump. I
opted instead for the most efficient natural gas furnace available from my
homebuilder, in
combination with the most efficient air-source
heat pump. Unlike GHPs, air-source heat pumps lack a ground
loop, meaning that they only work efficiently when temperatures are above about
40F. In my dual-fuel
system, the heat pump heats my house during milder weather (which is
frequent in Denver winters), and the natural gas furnace takes over when it is
cold. Since the heat pump is only slightly more expensive than the
air conditioner I would have bought anyway, the dual fuel system will pay for
itself rapidly, especially when natural gas prices are high.
From the perspective of the electric grid, my electric usage is higher and my
natural gas usage is lower during the heating season, when gas demand is high
and wind farms are at their most productive. So while a dual fuel
house is much less of a strain on the energy infrastructure than one with a
furnace and an air conditioner, it also saves the homeowner money for a much
smaller investment. In addition, while the need for a ground loop makes a
GHP nearly impossible to retrofit to an existing home, an air source heat pump
is an option for anyone considering replacing or installing an air conditioner,
and has the added advantage of having a back-up heat source during a natural gas
outage.
Another retrofit option I hope to see available soon is a hybrid
ground/air source heat pump [pdf]. These systems combine a short
ground loop with an air heat exchanger. By using the air exchanger during
milder weather, only a smaller ground source loop is needed for use during more extreme
conditions, reducing the up-front costs compared to a GHP, but without the
performance loss of an air source heat pump. A startup called Co-Energies
has developed a way to retrofit existing air conditioners into hybrid heat
pumps; see slides 33
and later of this PowerPoint.
Electricity Demand Can Shift
Heat pumps are just one option for changing the shape of the electricity
demand curve. Many such efficiency measures can do so. Other
examples are improved home sealing and insulation, which typically pay for
themselves in a couple years or less, and, because air conditioners work less
hard in the summer, reduce summer peak loads. Wind is undoubtedly a tricky
sort of electricity to use in the existing grid, but the fallacy that demand is
fixed makes the problem seem much harder than it needs to be.
