Blocking highs block wind turbine power

[David B. Benson]

This is a response to David Jones's comment on the BNC TCASE 15 thread --- Blocking highs don't happen very often; maybe with a recurrence time of 20 years or so here in the Pacific Northwest. Last autumn's blocking high did result in almost no wind derived power (I checked the BPA web page at the time) and the beginning did occur at the same time as the last portion of last summer's blocking high over Texas.

While there has been considerable talk about running HVDC transmission for long distances, most such fails to consider the high capital cost for relatively low average power transmitted, not to mention the transmission losses and that reliability is less than one might first expect.

Contrast with a new AC transmission line which might be finished later this year, from Boardman WA to Hemingway ID. Idaho Power will be the recipient of (only) 450 MW, which is likely to be nearly fully utilized immediately after completion. Idaho Power is quite short of reserves so using power generated by the Columbia dams, etc., further west and wheeled to southern Idaho over this new transmission line will enable Idaho Power to re-establish a more comfortable reserve. Their reserve requirement will only be fully met once the new CCGT under construction in southern Idaho is finished.

The Boardman to Hemingway transmission line is only about 500 km long and is taking three years to complete.

Summarizing the main point, nobody can afford to build reserves which are only to be used once each 20 years or so. As it proved during last summer's blocking high in Texas, ERCOT was down to about a 5% reserve, uncomfortably small. But at least that event in Texas did not result in rolling blackouts as has happened on two previous occasions fairly recently.

[dwj]

Reply to David Benson (Blocking Highs, May 9)
I thought for a while how to respond to this scenario and looked briefly at the situation in the Pacific Northwest. I looked for the wind lull you mention but could not see the period; I’ve presumably looked at the wrong dates – however lets accept that it occurred. I did find that the vast majority of the region’s wind farms are located in a concentrated area a bit smaller than metropolitan Melbourne. Because of this lack of geographical distribution I can well imagine the wind output would be very low on occasions. Because of the region’s large hydro resources this type of wind lull would not be a problem here, but let us look at a hypothetical region relying on wind and sun, with more typical hydro resources and project what could happen during a 6 week wind lull:
The hypothetical region has no interconnectors to adjacent regions and has the following power generation resources, expressed as a proportion of a nominal average annual load;

In: % of annual load energy provided, % capacity factor, Peak output as proportion of nominal load

Run of River Hydro 5%, 50%, 10%
Storage Hydro 5%, 20%, 25%
Biogas (from sewerage, municipal waste etc.) 2%, 100%, 2%
Other (user operated diesel, OCGTcogen, fuel cell etc.) 1%, 50%, 2%
Wind 43%, 30%, 143%
Solar 43%, 30%, 143%
OCGT 1%, 5%, 20%
Pumped Storage (10 days) 0%, Not applicable, 40%

The pumped storage would actually impose some load from losses but that simply forms part of the overall average load. The total rated peak capacity is 385%. This looks very high compared with a coal or nuclear based grid (which would be less than 200%) and it is, but the capital cost per peak watt is lower and it has very low running costs.

The scenario starts with the Pumped Storage at 80% capacity. Medium range wind forecasts predict the slow moving high pressure system 5 days ahead of its arrival and the OCGT are run up to full capacity to pre-store 1 additional “day” of energy. (In practice it would be a bit less than this because of competing solar energy storage requirements) During the event, output from the solar power schemes is increased due to the fine weather (this is a natural consequence of the persistent high pressure system). We assume a moderate increase of 30% over average (which is a reasonable assumption) giving an effective output of 57% of nominal load. The RoR hydro continues to be operated at average levels, 5% of nominal load, while the storage hydro can be operated at full capacity, 25% of nominal load. Biogas continues to operate at normal levels, 2% of nominal load, “other” generators are requested to run at full output, 2% of nominal load. Wind output is assumed to be only 1% of capacity, 1.5% of nominal load. OCGT can be operated at full output, 20% of nominal load. Pumped storage could run at up to 9/42 of stored energy per day or 21% of nominal load (with empty storage at the end of 42 days).

Summing all of these outputs gives 133.5% or in other words we do not need to run all of the generators this hard and could opt to reduce pumped or storage hydro output to reduce risk of shedding or we could reduce OCGT output to save fuel or various combinations. It would be a risk management decision as to how much OCGT use you trade off against running down the pumped storage. This scenario just shows how much flexibility you could have even in a once in twenty years crisis situation.

In a comparable nuclear based grid the hydro, biogas and “other” components would be unchanged. You would still need some peaking/standby OCGT reserve of similar capacity and it would probably be economical to have 10% to 20% pumped hydro storage (of lower energy capacity). Your nuclear or coal generators would need to provide 86% of the load energy and would need peak rating slightly above 100% of nominal load to allow for daily load cycling and downtime.

My apologies for the length of the post.

[David B. Benson]

dwj --- I am not yet ready to doubt the technical feasibility of your proposed solution to the blocking high problem. I do note you assume vastly more pumped hydro than is available in most localities around the globe. But more serious is than you didn't cost the system as opposed to an alternative with only a small portion of energy providing by wind. As wind turbines are, per unit energy, only slightly less expensive than NPPs, it isn't obvious that what you propose is more cost effective than something similar to the current French grid.

In any case, I'm only interested in low carbon generators so no coal burners and preferably almost no natgas burners (although using biogas and landfill gas is acceptable).

[davidj]

(Sorry, my previous persona seems to have been expunged)
David B, I didn’t provide any cost estimate as I was just intending to respond to the technical challenge you posed and to try to dispel some myths about renewables. You often see comments such as “every watt of wind needs to be backed up by a watt of gas generation” and “renewables can never provide baseload power”. These comments are simply untrue.

In my opinion a solar/wind based grid must have significant storage (and pumped hydro seems the only practical option we have at the moment) to account for the variability of both the generation and the load. Particularly that period in the late afternoon and early evening which is commonly peak load and where solar output is tailing off to nothing. This short period would often need pumped generation + normal storage hydro and even sometimes OCGT if the wind generation is low.

Solar schemes with a few hours of storage would assist greatly with this peak period. That seems to me the main benefit of thermal storage as it cannot negate the need for seasonal and poor weather storage – so you still need pumped hydro or something to provide bulk energy storage. (In which case what is the value of a few more hours of thermal storage?) Alternatively you just use more gas.

Costing of these options is a can of worms and depends on many assumptions which can bias the result in whichever direction you lean. A recent report for the Australian government (Australian Electricity Generation Technology Costs – Reference Case 2010; www.ret.gov.au/energy/Documents/AEGTC%202010.pdf) provides capital cost estimates as well as levelised cost of energy. Depending on actual technologies chosen, wind regime etc. it gives capital costs (in A$/kWpeak) of about:
Coal with CCS 5855
Nuclear- gen 3 5742
Wind 3577
Solar (central tower) 4559
OCGT 801
If you add in pumped hydro at about 4000 (my own guess) then the wind/solar scheme I described would cost almost twice as much as an equivalent nuclear grid (even more with transmission costs). However it would have running costs of about half that of the nuclear (and a third that of the coal with CCS) according to their numbers. (All of these numbers are significantly higher than the US costs which the report was based on.)

Personally, I have trouble swallowing their running cost estimates for nuclear as it specifically ignores waste and decommissioning costs and does not even mention risk (or insurance cost).