11 May 2011

David Keith on the APS Report

David Keith of the University of Calgary (pictured above) has posted up a detailed response to the APS air capture report (here in PDF).  Here is how it starts out:
The APS report does a fine job explaining the physics and chemistry of air capture (AC). We are pleased that the APS assembled such an excellent and diverse group of researchers to assess this technology at such an early stage in its development.

Our view about the costs of air capture diverges substantially from that of the APS. We acknowledge our bias and self-interest as developers of AC technology, but we believe that many of the key claims made here are independently verifiable.

The APS cost estimate for direct air capture depends on: (1) the performance of the APS reference design, (2) the choice of the reference design itself, and, (3) the validity of the cost estimating methods applied to the reference design.

We concur that the methodsused for cost estimation conform to industry practice, and applaud the report for laying these out in a transparent manner. In these slides we offer comments on items 1 and 2. In summary:

1.There are substantial technical discrepancies between APS’s performance model and estimates produced by industry-standard reference sources.

2.The choice of the APS’s reference design results in very high costs without a clear justification in terms of the tradeoff between cost and technical risk.
Keith also shares these comments with me by email, posted with his permission:
We have a paper under review that makes these points are more systematic way of lots of analysis.

Showing that a silly design is expensive proves nothing. To show that air capture expensive you need to show that a SENSIBLE design is expensive. That is, a sensible design at full industrial scale within a decade using low risk hardware.

An analogy: Suppose you are to estimating the cost of passenger jet traffic across the Atlantic in 1955. You need choose an aircraft design analogy as a starting point from which to scale and adjust your estimate. If you start with an F104 as your base case, one person per aircraft, you will get an foolish estimate about the actual cost per passenger of the 707 which was cheap because it was big and simple.

Starting with packed towers with stainless packing when there are commodity systems now on the market that will do the job for NaOH air capture is like starting with the F104. The APS did a fine job (ignoring the mistakes) costing the F104, but the choice of reference drives the answer.

Why did they not try several different systems and evaluate the cost-risk tradeoff? It's a mystery to me.
One thing here is for certain -- competition is good for both innovation and understanding.  Let's hope that it continues.

UPDATE 5/12: The following comments were shared with me by email for posting from Dr Tim Fox, Head of Energy and Environment, Institution of Mechanical Engineers, London, UK
I have read this report from the APS in full and its findings are based primarily on a detailed assessment of a single direct air capture technology proposal that has an energy intensive process at its core. The study also draws conclusions that largely ignore current thinking on the possible uses of this approach in climate change mitigation.

Air capture machines and processes represent a promising technology for the deployment of additional methods in tackling the challenge of global warming. However, given the early-stage and often proprietary nature of research and development in this area there is considerable uncertainty as to future cost levels. It is therefore premature to draw decisive conclusions as to the likely cost per tonne of CO2 captured. There are a number of other studies that suggest costs will be substantially lower than quoted by the APS and, indeed, given research, development and demonstration on a scale comparable with ‘conventional’ carbon capture and storage technologies (CCS) the approach might in certain circumstances gain near parity with that abatement option. We should however be clear that nobody is today seriously suggesting that in the case of power plant and large industrial emissions sources direct air capture would be considered as an alternative approach to CCS.

Air capture has the useful characteristic of being able to take advantage of the fact that direct removal of CO2 from the atmosphere can take place at any geographical location regardless of the point at which the gas is emitted. This enables the approach to be potentially used to tackle difficult sources, including large numbers of non-stationary emitters such as vehicles, aircraft and ships, or widely dispersed small scale industrial processes that are not amenable to CCS on technical or economic grounds. The technology also offers a wide range of other mitigation strategies including the removal of historic emissions accumulated in the atmosphere.

In summary, given the small amount of publicly available research and assessment data on proposed direct air capture technologies it is premature to draw any meaningful conclusions on the likely costs associated with the approach. As with CCS, the most promising technologies in this area need pilot testing and demonstration in the public domain to enable detailed cost assessments to be made, which will facilitate informed evidence-based decision making.

1 comment:

  1. Sometimes when you're dealing with a hairy problem, you need to step back and look at the totality of the challenge. So let's stop and think of all the emitters of CO2 in the world. All the individual car, truck and boat exhaust pipes, all the small motor exhausts, all the coal and oil and gas-fired business and factory and home chimneys.

    Now put them all in a row, scaled for output volume. Now you know the scale of the problem.

    All that CO2 produced is obtained for free - it is a by-product of profit or benefit-making processes. To capture the CO2 produced by all those emitters - for free - you need to start matching them in number and scale. And you will not be doing it for free. This time, you pay to reverse what the emitters have done as a by-product.

    Now that you have some context, the question becomes, just how low would you have to bring the cost down to even start matching the emitters? Remember - these CO2 removers don't produce a product you can sell. They run at a pure loss of money. And you're going to have to run them at that loss for the life of the machine.

    It seems to me that in that context, carbon capture is less daunting technical task and more fool's errand. For instance, what would it cost to build and run the equivalent of a small machine for every car, truck and boat on the planet?

    If people want to use their own money answering these questions, then have at it. But if it's tax money you're talking about, then opportunity cost rears its ugly head.