A Purely Cosmetic Gesture?
A cynic might observe that the rules mean nothing, because new coal-fired power plants in the US were already pretty much a dead issue. Even without CCS, the capital cost of a new coal-fired plant is several times that of a natural gas combined cycle (NGCC) plant. The higher capital cost for coal has traditionally been balanced by lower fuel costs. Recently, however, bargain prices for natural gas have largely erased even that advantage. Utilities needing new capacity have therefore been dropping plans for pulverized coal (PC) or nuclear plants in favor of NGCC. As a practical matter, therefore, the EPA rules for new power plants are a symbolic offering to climate activists that will have no real impact on the power industry.
I don't entirely agree with the cynics. Even with the proposed limit on CO2 emissions, I don't believe that coal is a dead option. Nobody really trusts that the bargain prices we're currently seeing for natural gas will hold for long. Natural gas has traditionally been less costly than oil, in terms of dollars per Btu, but the disparity now is unprecedented. At current prices, long-haul trucks could slash their fuel bills by two thirds, if they could convert to natural gas and if the necessary refueling infrastructure were in place. But both oil and gas are supply-limited commodities. Production runs near the limit that the existing set of production wells can manage. With little idle capacity, it isn't possible to simply dial up production when price signals indicate increasing demand. Small increases in demand can therefore lead to soaring prices. Awareness of that fact has so far made players cautious about switching.
Coal, by contrast, is still a demand-limited commodity. There's plenty of coal available for mining, whenever prices get high enough to justify the effort. Stockpiling is easy, and prices are much less volatile. Maintaining some level of coal-fired generating capacity is a therefore a reasonable hedge for utilities to protect themselves and their customers from a sharp rise in natural gas prices.
If I'm right, then we will (eventually) see new coal-fired plants being built, as well as major upgrades and conversions of existing plants. In that case, the EPA's proposed greenhouse gas rules -- assuming they go into effect -- will significantly affect the design of those plants and conversions. They will make some designs more attractive and others less so.
Designs that will be less attractive include the current generation of conventional PC steam power plants with their "bag house" filters and flue gas scrubbers for control of emissions. While it's possible to integrate another level of flue gas scrubbing to remove CO2, the result is not appealing. The IPCC, in its 2005 Special Report on Carbon Dioxide Capture and Sequestration, estimated the cost of a new PC plant of that type at about $2100 per kilowatt (vs. ~$1300 without CCS)1. Those specific dollar values have become dated by rising commodity prices since 2005, but they still give an idea of the high relative cost for post-combustion CCS.
It's not just the capital cost that is affected. Operation of the CCS subsystems would constitute a parasitic drain of around 24% on the plant's power output. That breaks down to 13% for steam to regenerate the CO2 sorbent, 9% to compress the CO2 to 100 atmospheres for transport and sequestration, and 2% for other pumps and blowers2. A 24% parasitic drain means a 32% increase in coal consumption per MWh delivered. That's pretty bad. Nevertheless, it represents a proven "baseline" design that could be built today, and is the starting point for most discussions of CCS.
It is almost certainly possible to do better. Some companies are investigating a more efficient scrubbing technology that uses chilled ammonia instead of the amine-based solutions currently employed. Another promising alternative employs "oxy-fuel" combustion in the boiler.
R&D on oxy-fuel combustion is mostly being carried out in Europe. Instead of air, the pulverized coal is burned in an atmosphere of re-circulated CO2 with oxygen injected. The resulting flue gas is over 90% CO2, and its volume is about one fifth what it would be with air combustion. That raises boiler efficiency, since less heat is carried off in hot flue gases. It also makes it easier to precipitate fly ash and other particulates. But most importantly, the concentration of CO2 is high enough that after relatively simple processing to remove water vapor, the gas can be piped directly to the sequestration site. It is not released into the atmosphere, and the power plant has essentially zero emissions.
The down side of oxy-fuel combustion is the need for an associated air separation unit (ASU) to produce the oxygen. However the cost and parasitic energy draw of the ASU are projected to be lower than they are for the amine-based scrubbing units of the baseline CCS approach. Reference2 puts the capital investment premium for an oxy-fuel plant (over that of a conventional PC plant without CCS) as 14% (vs. 23% for the post-combustion CCS approach), and the parasitic energy drain as 20% net (vs. 24% for post-combustion CCS). That still makes CCS a costly option to support, but with the oxy-fuel approach, the premium looks to be only about two thirds of what it would be for the baseline approach.
It is likely that in the future the premium for oxy-fuel combustion can be further reduced by new air separation technology. Researchers from Praxair, working with the University of Utah, have been investigating ceramic "oxygen transport membranes" (OTMs) for separation of oxygen from air at high temperatures. According to a presentation to DOE in 20093, integration of an OTM-based ASU into the boiler has the potential to reduce the parasitic energy cost of oxy-fuel combustion by 75%. The technology won't be available tomorrow; commercial deployment is not projected to begin until the 2020 time frame. But if the technology lives up to its developers hopes -- always a big if -- it should be able to produce power with CCS at a small cost premium of 10% or less compared to a conventional plant without CCS.
What about IGCC?
A cautionary tale about "developers hopes" can be found in the progress of another alternative coal technology: "integrated gasification combined cycle" or IGCC. In that approach, coal is gasified to produce a combustible coal gas. The gas is cleaned of ash and sulfur, and then fuels a gas turbine combined cycle power plant. The original motivation for the process was not carbon sequestration, but higher thermal efficiency and the ability to consume high sulfur coal. By its nature, however, the process can be modified to produce separate and relatively pure streams of CO2 and hydrogen at little additional cost. At least, that's the theory.
In practice, IGCC has not fared so well. The capital cost is high, but that was expected. What was not expected was the difficulty in keeping the gasification process running smoothly. It has proven "fickle" and sensitive to the type of coal being consumed. Pilot plants have experienced extended down times for costly maintenance and repairs4. Once viewed as the flagship for "clean coal" technology, IGCC's promise has been tarnished. It still has its advocates, and the DOE has an ongoing funding program for research into improvements. It may yet achieve its promise, but the improvements being studied suggest something of a "back to the drawing board" retrenchment5.
Better Turbine Technology
There is one further technology that I feel should be mentioned here. It's not a technology for carbon capture, per se, but it could have a major impact on the future of coal-fired power plants in general -- and therefore on the future of CCS under the new EPA rules. It's a technology that has been under study at Sandia Labs, nominally for enhanced performance of upgraded nuclear power plants. Nothing about it is specific to nuclear power, however, and it has generated broad interest for use with concentrated solar platforms as well as a new generation of coal-fired power plants. It is the supercritical CO2 (S-CO2) Brayton cycle turbine6.
The S-CO2 Brayton cycle has two key characteristics that have been touted in articles about it:
Potential for Throttling?
A third potential characteristic of this type of turbine is one that I have not seen mentioned anywhere. It is speculative on my part, but I believe the turbine could be designed to be efficiently throttled over a fairly wide range of power outputs. That's because the high density of its working fluid enables it to use a single radial compressor in the cold gas compression section and a cascade of only three radial expander stages in the hot gas expansion section. The stages resemble those of a hydroelectric turbine, rather than the many stacked disks of rotor and stator blades in an axial flow turbine. Why is that significant?
The answer is that in a radial flow arrangement, compression or expansion are primarily due to centrifugal force. They are largely independent of the flow rate of the working fluid. The power developed by a hydroelectric turbine, for example, can be varied by controlling the water flow rate through an adjustable headstock nozzle. That's why local hydroelectric generators are highly valued by transmission system operators; their power can be quickly ramped up or down under real-time control for load balancing and grid stabilization. It seems to me that it should be possible to design a supercritical CO2 Brayton cycle turbine for the same type of regulated operation. If so, it would make coal-fired power plants of this type, equipped with CCS, the perfect choice for backing intermittent wind and solar power.
So what conclusions can we draw? We've looked at the capture side of CCS, but not the sequestration side. The options available for sequestration do have significant bearing on the feasibility and economics of CCS, but I'll save that discussion for later, in part 2. The technology picture described above should nonetheless be enough to support some tentative conclusions. To sum up:
In part 2 next week, we'll look at the sequestration side of CCS. Then in part 3, I'll step back a bit and try to examine some of the broader issues around CO2 emissions in general.
Notes & References