Press reports in the Financial Times and other news outlets describe a wind project in Oregon with 338 machines of 2.5 MW each, giving a total capacity of 845 MW. The project sponsors claim that they will provide enough energy to serve 235,000 households and reduce CO2 output by 1.5 million tonnes annually.
Part I demonstrated that the served-household claims is fanciful. In reality, no more than 49,000 households could be “supplied”, and these with only a minimal degree of assurance. Indeed, the wind project is more costly than a diesel backup scheme that would actually be capable of supplying reliable power to several hundred thousand households. The wind project is also three times more costly than a replacement of just 211 MW of older coal capacity with new technology that would provide a similar reduction in emissions, while supplying firm power to the NW Power Pool’s customers.
Opportunity-cost economics, anyone?
The key to wind’s providing some degree of fuel and emissions savings is its ability to deliver reliable electricity without shadowing or backup by hydrocarbon-using plants. These shadowing/backup requirements in the Northwest (NW) Power Pool may be able to take advantage of existing surplus hydro capacity in that region during off-peak periods (spring and fall), thereby permitting the proposed plant to reduce hydrocarbon consumption and emissions somewhat during those periods. It is not reasonable to expect to achieve the claimed emissions savings, but lower figures, less than half the publicized savings, may be possible.
In particular, the addition of wind generation, with shadowing/ backup provided by reservoir hydro, may be able to reduce overall CO2 emissions in California, the ultimate customer for the electricity produced by the GE project during Oregon’s two surplus seasons. But during the winter and summer peak demand periods, less hydro output is available, peak demand is greater and the shadowing backup will be provided by some combination of gas-fired and coal plants. What it is critical to keep in mind is that maintaining stability in the NW Power Pool requires the pool to shadow/backup not only the proposed new project, but the other 6.4 GW of existing wind as well.
Going further, our analysis shows there are less costly and more effective alternatives readily available that rival or exceed the claimed benefits of this wind project.
Wind Shadowing/Backup Requirements
So what is needed to ensure that wind plants deliver reliable electricity? They have to be paired with conventional, reliable generators capable of mirroring wind’s volatile and unreliable output. This can be called wind shadowing/backup capacity. It is shadowing wind when wind is producing, albeit it in a volatile manner. It is backup to wind, in the more usual sense of the word, when wind is producing nothing, which can be for extended periods.
When claims are made about wind displacing fossil fuel plant production, the question that should be asked first is: what is providing wind shadowing/backup? With system reliability and power quality considerations coming to the fore, it becomes evident that the shadowing/backup is what is displacing the fossil fuel production, and wind is displacing some small measure of the shadowing/backup. An earlier article explored the realities of this and showed that a wind project that relies on fossil generators to shadow the wind machines may provide little net fuel or CO2 displacement and in some cases may actually increase fuel use and emissions. The latter result may obtain as a result of: (1) the imposed inefficient operation of the wind shadowing/backup, as well as (2) use of shadowing/backup technologies that are less efficient than the pool’s major generation resources – coal, nuclear, gas-fired combined cycle. The three generation sources listed above are in varying ways not generally suitable for providing shadowing for wind. In each case the ramp rate of the generator is too slow in reacting to many of the transients of wind production. Consequently, shadowing and backup must be provided by smaller, faster acting, but less efficient engines. If the shadowing/backup requirements are significant – that is, if wind output is large relative to overall system capacity, even approaching 5% – then the reliance on small, inefficient engines or combustion turbines (GTs) will arguably lead to a net increase in fuel use and therefore emissions.
The general considerations are:
Normally, the full and accurate computation of the technologies involved in shadowing/backup of wind will require a system dispatch model so that minute-by-minute variations in wind output can be shadowed by fast ramping engines or valves (hydro). Table 1 summarizes some of the possible scenarios.
Table 1 – Some Wind Shadowing/Backup Scenarios In the NW Power Pool
Gas turbine (CCGT, OCGT)
Fall (relative to coal)
Rise (relative to CCGT alone)
|Gas turbine (CCGT, OCGT)
|Gas turbine (CCGT)
|Hydro (run of river)
|Hydro (run of river)
|Hydro (run of river)
In scenarios A, B and C, the inefficiencies imposed by wind volatility on the shadowing/backup plants can more than offset the CO2 emissions “saved” at the point of wind generation. In any event, Scenario C is relevant in the NW Power Pool only insofar as coal is used as a resource in the pool, and coal-fired electricity enters this pool largely through imports. In case D, assuming no curtailment of wind during high wind production periods and no spillage of hydro is required because of the timing of wind production relative to reservoir levels, the wind production could be replacing that of fossil fuel, as indicated by “Other”. In case E, wind is replacing hydro and no CO2 emissions are saved (generally wind acts similarly to run of river hydro, in terms of system stability, with the exception of such cases as hydro plants at Niagara). Note that the conditions for case D are seldom met during annual peak demand periods in the NW Power Pool, as noted in Part 1.
The Oregon wind plant production is slated to go to Southern California Edison, which obtains over 50 per cent of its electricity from imports (out of state) and almost 40 per cent from thermal generation within its jurisdiction. As California as a whole gets 50 per cent of its in-state generation from natural gas and about 2 per cent from coal/oil, it is reasonably assumed that the wind/shadowing-backup combination is displacing gas, mostly in combined cycle plants. It is possible that some imported electricity is being displaced, which likely contains a higher proportion of coal.
The question remains: what is being used as wind shadowing/backup? Oregon has the following electricity production profile – hydro 61 percent, gas 27 percent, coal/oil 8 percent, and other renewables 4 percent. A reasonable assumption is that impounded hydro is being used within Oregon for this purpose during shoulder seasons (spring and fall), while gas and possibly coal are used during peak seasons (Summer and Winter). In off-peak seasons in Oregon and the NW Power Pool, case D generally applies and Oregon is basically exporting hydro and some wind. Case A or B applies during peak seasons, and gas or coal is likely exported.
CO2 Emissions Saved From Wind Generation
The foregoing illustrates the complexity of determining the impact of wind plants on fossil fuel and CO2 emissions reductions in electricity systems. The following completes the application of this to the new Oregon wind plant.
The wind project sponsors claim that 1.5 million tons of CO2 emissions per year will be saved as a result of this investment. Accepting the premise that no shadowing/backup will be needed the most likely result is for the wind to displace gas-fired CCGTs, at 0.4 tons CO2 emissions per MWh:
845 x 0.30 x 24 x 365 x 0.40 = 890,000 tonnes or about 0.9 million tonnes per year
For a 25 per cent capacity factor, more reasonable for onshore facilities, the CO2 emissions saved become about 0.7 million tonnes per year. The actual savings are likely to be far less than this calculated figure, since hydro capability is reduced during the winter peak demand period, one that coincides with troughs in wind availability as well. As a result, and as indicated above, the NW Power Pool is likely to be exporting gas/coal generated electricity to Southern California during the winter demand peak as well as during the summer peak. In fact, any coal-generated electricity exported to cover the supply obligation of the wind farm is likely to come from the same plants in Utah, Montana, Arizona and Nevada that currently provide the overall grid stability for Southern California Edison and California in general – a contractual round trip that contributes little or nothing to net energy supplies and saves little or no fuel/emissions.
It should be noted that potential savings of fuel/emissions during shoulder periods (fall and spring) comprise a special case because of the large hydro capability in Oregon during such periods. In the more general case and during the summer and winter peak demand periods, with gas or coal used for wind shadowing/backup, the CO2 emissions savings would reverse and net fuel use/emissions would rise due to the inefficiencies imposed on these plants. In fact with wind, currently at 6.4 GW, expected to approach 10% of pool generation capability in the NW Power Pool with the new project, the ability of the smaller, faster responding and more efficient shadowing engines described in Power Magazine are likely to be impracticable, since more than 800 of such engines would be required, meaning that shadowing/backup will be supplied by gas turbines, with the attendant inefficiencies and high fuel consumption, especially during startup. During lulls in wind a system of this size will require significant conventional generation resources for shadowing/backup.
At this point, a reasonable expectation is that half of the reduced CO2 emissions shown above would be achieved, given that the generation savings are valid for roughly half the year, spring and fall seasons; that is:
50% of 0.7 = 0.35 million tonnes per year
Since providing shadowing/backup for the NW Power Pool’s overall wind generation capacity of 7.3 GW, including the proposed GE project, involves large combustion turbines, then the fuel used just for startup, about 8-10 tonnes for each turbine each time, needs to be debited from the emissions reductions account to the wind plant. Each startup cycle, using liquid fuel or pressurized gas, produces about 100 tonnes of CO2 . To back up the NW Power Pool’s wind capacity would put roughly an additional 90 million tonnes of CO2 into the air, that is:
75 units x 12 start ups for each x 100 tonnes CO2/startup = 0.090 million tonnes per year
The GE project’s share of the shadowing/backup startup CO2 emissions (~11%) would be roughly 9,900 tonnes, offsetting about 3% of entire calculated CO2 savings for the 845 MW project.
Emissions savings identical to those claimed for the new wind project can be accomplished at significantly lower cost simply by replacing older coal-fired power plants (<35% conversion efficiency and relatively dirty) with current “ordinary” coal fired plants (~41% efficient and much cleaner). “Ordinary” current technology would reduce emissions in the pool by 0.22 million tonnes/year for 211 MW of firm capacity, roughly the amount of energy that the proposed wind project generates. Higher technology coal plants (~45% efficient and very clean), more efficient still, will reduce emissions by more than 0.35 million tonnes/year for the same amount of electricity generated by 845 MW of wind. As noted previously on this blog, many willing investors are anxious to make such investments. Only a perverse system of government permits and approvals and uninformed environmental groups stands between newer combustion technology and improved power supply. These are truly the “shovel-ready” projects.
The costs to the electricity consuming public for emissions reductions on the order of what is produced by the proposed Oregon wind plant are less than one half what will be required to keep the new wind project in operation and shadowed/backed up properly. An investment of a similar magnitude to the wind plant in high technology coal combustion, by replacing roughly 1,000 MW of older, less efficient, dirty coal generation capacity, would reduce emissions of CO2 (and a lot of other things like SOx and NOxand mercury) by more than 1.65 million tonnes annually, more than five times the emissions reductions that can be credited to the wind plants with the plus of a substantial improvement in grid reliability. Investing $1.9 billion in new high efficiency coal plant of 845 MW could replace older ones and reduce emissions considerably. Alternatively, such a plant would serve 600,000 additional household customers in the NW Power Pool or Southern California for about 6.5 cents/kWh, roughly one third the cost of wind, including its shadowing/backup requirements without the need to resort to arithmetic sleights-of-hand about reliability.
The considerations of wind availability, system operations and hydro availability are likely to be more complex than the treatment given here. However, a more complete system simulation is unlikely to be more favorable to wind than is the present treatment, especially if increased reliability standards are implemented for power pools. The proposed CO2 savings from the Oregon wind project are overstated to a significant degree and it is likely that net fuel/emissions savings will only be possible during periods of surplus hydro availability – the off-peak spring and fall seasons.
The lesson from this case is that reported claims of benefits from the introduction of industrial wind plants, such as, households served and CO2 emissions saved should be carefully reviewed – they are generally difficult to support.
And since wind competes with other projects for investment capital the funds that are devoted to wind may actually reduce potential emissions savings from efficiency and technology improvements in coal, improvements that can be supplied without tax credits or other fiscal chicanery.