The Jacobson Roadmap

Figure 1. Bonneville looking east. The new dam is to the left of the spillway. The bit of land between the spillway and the new dam is the original north bank of the river.

After my disastrous foray into reporting breaking news, it's time to quickly retreat back into sermonizing and hope the choir has very short memories.

Mark Jacobson's 2015 claim that an all wind/solar/water grid was economically feasible for the US was almost immediately debunked by Clack et al. I figured it was long dead issue. I was wrong. Seems to be my standard mode these days. It has recently been used to justify the New York's Climate Action Council's decision that New York did not need any new technologies to handle periods of low wind and solar to meet New York's impossible net zero targets. This old analysis addresses only one of Jacobson's implausible claims.

Because of their intermittency, wind and solar are supplemental sources. There are situations where these supplemental sources can make good sense. Maui is probably a case. Strong, relatively dependable winds, lots of sun and the alternative is diesel with a fuel cost of 20 cents/kWh. Wind/sun can make economic sense in such places; but what you must do is compare the fully built up cost of the intermittent source with the fuel cost of the dispatchable source, because the only thing the grid saves is the fuel that would have been burned if the W/S capacity were not there.

Another situation in which a supplemental source can make economic sense is in places which are blessed with both lots of wind or sun and lots of hydro. Hydroelectric plants are usually designed to handle periods in which the river flow is high. During low flow periods, the available capacity is less than the nameplate capacity. Wind and solar can take some of the pressure off the hydro capacity allowing the reservoirs to build up and increase their effective dispatchable capacity. The windy Columbia River is a place where this combination can work. But few places meet this criteria and even there the impact is marginal.

In a 2015 paper, Jacobson made the astounding claim that US hydro "instantaneous" capacity could be increased by a factor of 16 in this manner.\cite{jacobson-2015} Figure 2 shows a portion of Jacobson's simulation of his all wind/water/solar US grid. The peak demand on hydro is 1300 GW's, and we need that capacity for 12 hours. And we need about 800 GW's 12 hours later for another 12 hours. And so on. Currently, the US peak hydro capacity is 79 GW.

Figure 2, Simulation of Jacobson grid, reference \cite{clack-2017}[Figure 1].}

Jacobson implied that the ``instantaneous" discharge capacity is far higher than the nameplate capacity. But ``instantaneous" in this context is something like 12 hours. The single biggest source of hydroelectric power in the US is the Columbia River. I live on the Columbia a few miles upstream from the Bonneville Dam, Figure 1. The original dam finished in 1938 has a nameplate capacity of 518 MW with an overload (aka instantaneous) capacity of 574 MW. The original dam was undersized for the river flow. In 1982 a ``new" dam and powerhouse was completed. Its nameplate capacity is 532 MW; overload is 612 MW.

Jacobson needs to increase Bonneville's instantaneous capacity by more than a factor of ten. But there is no room for more generators. Except for the necessary spillway, the powerhouses extend all the way across the river. In fact, when they built the "new" dam in 1982, they had to widen the river. The dug a 1/4 mile into the north bank to make room for the new powerhouse. In doing so, they had to displace some of the people they had already displaced in 1938.

The current overload capacity of Bonneville is 1130 MW. When Bonneville is going all out in late summer, the river falls like a rock, at least a meter per day. With ten times as many turbines, the river will drop 10 times faster. And as the river drops, Bonneville puts out less power. The only way you can materially increase the 12 hour discharge capacity is to not only install a whole bunch of new turbines, for which there is no room, you must build a higher dam.¹

You also need to inundate the railroads on either side of the river, an Interstate, and a whole series of river towns. The original dams took the water level up to something like the 50 year flood level. People had responded to the floods by not building a lot of stuff including the railroads below those levels. Yet there were still severe dislocations when the current dams were built. To go higher would start a war. And that only gets you through one day.

The Columbia is pretty much maxed out. And the same thing is true of most first world hydro resources.

1. This will not get you more vertical. Except for one short stretch, the Columbia is fully dammed. Increasing the height of the Bonneville Dam decreases the effective height of the next dam upriver at The Dalles, and so on.