18 Comments

Summary:

In the grand scheme of energy technologies, the key component that makes up a fuel cell is relatively short-lived. This Achilles heel is one of the main reasons that the business of selling fuel cells is often times not profitable.

AdobeBloom2

In the grand scheme of energy technologies, the key component that makes up a fuel cell — which is like a chemical battery that produces electricity — is relatively short-lived. This Achilles heel is one of the main reasons that building, installing and selling fuel cells can be so expensive, and almost none of the fuel cell makers are profitable yet.

Of course, there are different types of fuel cells, but in general, the stacks that make up a fuel cell, and create the reaction that produces electricity, often last only about two to five years. This is common for different types of fuel cells like solid oxide fuel cells (Bloom Energy makes this type) or proton exchange membrane (PEM) fuel cells, like what ClearEdge Power builds.

A fuel cell’s stacks fill a chamber called the hot box, and it’s this chamber that gets swapped out of these fuel cells every few years.  The stack contains a catalyst, often platinum, which, when combined with the fuel source (natural gas or hydrogen) and oxygen create electricity.

Break down

Over time, as the fuel and oxygen are constantly being pumped in and run over the catalyst in the stacks, the chemicals start to degrade and the system starts to wear down.

Fuel cells are similar to a battery in their degrading process (see Why lithium-ion batteries die so young), and fuel cell stacks, like a battery, have an anode and cathode portions. Fuel cells also run at high temperatures, which is another reason these systems degrade quickly.

The short life span of the hot box is a key problem for the capital costs of fuel cell makers. The hot box can make up a significant portion of the fuel cell, and I’ve heard as high as 50 to 75 percent of the cost of the system. That cost can be lower, however, and for example, ClearEdge Power’s VP of Marketing Mike Upp told me the stacks in a ClearEdge fuel cell can make up 25 to 30 percent of the cost of the system.

Costs climb

Fuel cell makers are toiling away at trying to extend the life time of the hot box, as well as reduce overall manufacturing costs. Upp said that while the stacks in ClearEdge’s first iteration of its fuel cell last three to five years, the company’s engineers are working on doubling and tripling that lifetime every few years. Stacks can also be recycled, which can reduce the overall capital costs.

Fuel cell makers are spending a lot on R&D trying to find these stack lifetime breakthroughs, but are also looking to reduce costs via reaching economies of scale of manufacturing. The idea is even if the stacks don’t last longer in the future, they can ultimately be cheaper to produce. Bloom Energy has been scaling up manufacturing of its solid oxide fuel cells, and NEA Partner Scott Sandell told me back when Bloom launched that it would be the economies of scale that would push down costs dramatically over the years.

I heard a rumor recently that Bloom Energy had closed yet another round of $150 million in funding, which would bring its funding raised to over $550 million. Earlier this year, VentureWire reported that Bloom had quietly raised about $100 million more in equity, above its confirmed $400 million. No doubt part of these funds are going to both R&D to extend the life of the hot box, as well as the capital costs to actually replace the hot boxes for its first customers.

We’ll see if any of the leading fuel cell makers can effectively reduce their costs enough, and lengthen the lifetime of the hot box. If they are successful with that, then more of these companies could be profitable one day.

You’re subscribed! If you like, you can update your settings

  1. Technology improvements are extending stack life dramatically over the last 10 years and are expected to continue to increase with more operation. The biggest thing that residential type users do not realize is the demand on a power generation system like this. There is no other household appliance with such a demand. Turn your computer on and run it at capacity for 5 years continuously and see if there is not a component that needs to be replaced. Oh, computer lifecycles are less than 5 years. People will say refrigerators (only runs 5-10 minutes per hour) or there car (maybe 2-3 hours per day) have the same demands, but they do not see they do not run all the time. Stationary FC systems are made to run continuously. Few other devices outside of electrical generating systems and satellites have that kind of demand. Give other examples of household equipment with a 40,000 plus life expectancy. Even an LED light (relatively low temp, no chemical reactions, no issues with air quality, water quality or other environmental effects) has a life expectancy of 50,000 hours.

    1. You raise some good points:
      Fuel cells are meant to run continuously (and ideally at capacity). This reveals one of the core problems with fuel cells – they don’t follow load particularly well. I haven’t seen any detailed cut-sheets from Bloom, but other manufacturers (like UTC or FCE) do provide them. Low max steps from one steady state to another and hefty recovery times limit the applications that a fuel cell can fill.

      And on power gen equipment lifetime between major service…you neglected to mention recip engines or turbines, both of which – run continuously – have longer MTBMs than fuel cells.

  2. When compared to the R&D being done on Li-on or even Li-air batteries,research on fuel cells seems to be much less.

    1. Maybe that’s because most people who actually understand fuels don’t really believe they can ever be economically viable. Marketing of these products hinges on buyer ignorance.

    2. Look at the dollars being invested in fuel cells compared to batteries today–you’ll see why. As long as dollars pour into universities and national labs to focus on batteries, they will, and they’ll continue to churn out papers and self-promotion. Look back five to eight years, and you’d see the same with PEM fuel cells.

      On another note, the media has a real difficulty separating solid oxide and PEM fuel cells, and this is a shame. They love to wield the platinum paintbrush as a hurdle to fuel cells when solid oxide fuel cells don’t contain platinum at all.

      There continue to be hurdles for fuel cells and batteries, and they both deserve funding. It’s a shame that even with a award winning scientist as the Secretary of Energy, the U.S. can’t develop a balanced research program, but oscillates between silver-bullet answers.

      All this said, the warchest that Bloom has amassed is formidable–they have a product and they are working to reduce its cost. The materials set in their stack is credible and they appear to be extremely innovative in strategies to bring the technology to market. If they can be successful, and their stockholders win, it could break a logjam for funding all energy technologies. Let’s hope they succeed.

  3. Katie Fehrenbacher Monday, August 8, 2011

    @James J, @darkhorse, insightful comments, thanks for these.

    1. I have worked in both the semiconductor and fuel cell industries. Early semiconductors did not have predictable lifetimes. I participated in quality and reliability research which has lead to the semiconductor reliability we have today. A relentless focus on materials research and manufacturing quality ignited by the Deming principles was required. It was a culture change, on both sides of the Pacific.

      The fuel cell industry requires the same management focus. It’s there too, in many companies. For example there is strong Japanese participation in both the stationary and automobile fuel cell industries by companies with a modern quality and reliability culture, and staffing to match.

      There are no fundamental material limits to stack and balance of plant longevity.

      The fuel cell industry is in very early volume stages which is documented in the DOE 2010 fuel cell survey. The square meters manufactured yearly worldwide, of say GoreTex, dwarf PEM material; it would be similar for SOFC substrates and materials in other fuel cell technologies. Volume is inversely coupled to cost.

      The value of fuel cells is their ability to extract energy from chemical bonds very efficiently, with few unwanted pollutants, and silently. The systems can span peaking response as does Ballard’s 1MW system in utility trials, and base load, for example, the systems mentioned.

      The industry would be better served by focusing American, Japanese, European and Korean research efforts on reliability engineering, shared worldwide, and on shared failure analysis facilities.

  4. Fuel Cells have been trying to get stack rebuild costs under control for decades.

    This is the Achilles heel for sure, ans most are based on both precious and scarce materials for catalysts, and all the precious metal prices have gone way UP by 3-4X, so I really don’t see why they are even financially viable anymore.

    ALL the efficiency/environmental benefits of fuel cells are more than LOST with JUST the stack rebuild cost differences vs. turbines and recips when amortized over their operating life/kwh.

    34% vs. 48-52%? Meaningless if your Bloombox costs upward of $400K/100kw to fix. $400,000/80000hrs (I think I’m being generous here as I’m guessing the stack life can vary between 3-5 years)/100kw is a whopping 5 CENTS/kwh. An engine can be rebuilt for a fraction of that cost (about a penny or less) amortized over the same term. The rebuild costs of other fuel cells, while less for some (UTC), still wipe out ANY efficiency/revenue gain if the true cost differences in O&M vs. an engine or turbine are properly SUBTRACTED from the efficiency/revenue ‘gains’ of the fuel cell.

    All in all, it looks to me like a very expensive zero sum game and a high priced alternative to more mature, stable, lower risk and off the shelf technologies available today at a fraction of the cost to both procure (with or without Government Monopoly Money) and maintain.

    And, even if you were to get the first Fuel Cell for free, you’re STILL on the hook for a FAT O&M agreement that can chew through your ROI like a ravenous parasite.

    Now lets talk about life cycle costs, and what special toxic disposal procedures might need to be followed when recycling a stack with high concentrations of toxic/precious metals in them…as opposed to engines that can simply be scrapped and melted down?

    There go the environmental benefits.

    Kind of like having to recycle spent fuel rods from a nuclear plant, after years of ‘clean’ energy production….except with fuel cells, it can be done, at some additional expense to the environment…and the pocketbook, I suppose.

    Bottom line is that you can’t just look at the ‘benefits’ of a technology without looking at the total big picture life cycle costs.

    To neglect to do so does no service to the customer, the pocketbook or the environment.

    1. Saying that fuel cell technology has an Achilles heel is an understatement. It’s a whole centipede of problems. Nothing works right – not hydrogen generation, not transportation, not storage, not conversion into electricity. It’s good pipe dream though.

    2. All very good points, but….
      you don’t give the recip engines and turbines enough credit. 34% efficiency? A bit too harsh.

      34% is on the low end of anything in the Jenbacher lineup. The larger 4 and 6 series engines pull down 42-44% efficiency and the newer 9 series is looking to be mid-to-high 40s. Wartsilla’s been working in the mid/high 40s for its larger gensets for a while now.

      Turbines are a tougher sell, but recouperator models like the Solar Merc 50s can generate at 38-39%. But you wouldn’t really install a gas turbine for electric efficiency, then would you?

      And on the flip side, 48%-52% is a bit too ambitious…even the Bloom units are topping out at 48 or so….and they don’t get *any* process heat – it all goes right back into the reformer. And the Purecell’s aren’t doing much better than 40%. FCE’s 300kW model is a tad better at 42, IIRC.

      All of that said, your point remains about the O&M and costs stack replacement costs.

  5. Several of the solid oxide fuel cell developers have been in a Federal Government $billion hardware cost lowering program with the government called the SECA, now for several years. One of these, Versa Power has lowered, not its stack costs, but its complete fuel cell costs to only $700 per kW under volume production and anticipates that in 2012 its costs will be only $400 per kW. Its stack costs are only a part of that, likely only one third to one half.
    Before the financial crunch, Versa Power was looking for money to build a commercial factory — it currently operates a pilot plant in Canada built by Global Thermoelectric. Global sold out to Fuel Cell Energy. FCE spun off all its tangible plant to Versa Power that had been jointly started by EPRI the electric industry’s research organization and by a similar organization for the natural gas industry. The DOE issues progress reports for SECA from time to time. They make good reading.

  6. SECA stack cost in 2010, $175/kW.

  7. Katie – I think you have to be careful with statements, such as “which is like a chemical battery that produces electricity.” A fuel cell is really an energy conversion device, so a better analog, in some ways, is a gas turbine (which also had to go through its own high-temperature materials development phase, and still is as turbine temps go high and less pure fuels, e.g., syngas, are being contemplated.

    In SOFCs the challenge is trying to match up dissimilar materials with different expansion rates, oxidation rates, etc. Keeping the units sealed and the interconnects working are tough engineering questions. One research direction is to drive operating temperatures down, so that cheaper and somewhat less durable materials can be used. The other research direction is to come up with more durable high-temp materials and coatings (requiring a mix of alloy, glass and ceramic materials wizardry). Progress, even slowly, is being made. For example, I know that one well known supplier to SOFC makers, NexTech, said today it has a developed a ceramic coating (MCO) that can be used with low-cost steel interconnects that have been tested with a service life of 40,000 hours at 750°C. That is still 4.5 years of continuous service, but the cheaper steel is an example of lowering the replacement costs.

  8. It is also a point of interest to note the potential hydrogen production industry for fuel cells. Currently H2Sonics (working out of Proton Onsite’s factory in CT.) has won the patent rights for a portable, scalable hydrogen generator that produces both hydrogen and ultrapure alumina in a compact-sized unit without the use of fossil fuels or environmentally hazardous material. If this product becomes viable on a production scale then hydrogen prices will drop making devices that utilize hydrogen (such as fuel cells) a more economic alternative as the aluminum market is already mature.

    http://www.h2sonics.com/

    http://www.protononsite.com/

  9. Fuel cells using syngas are more than contemplated. They are even now being used to construct a megawatt class fuel cell system for DOE. DOE has a contract with Fuel Cell Energy Co. to build it with its affiliate Versa Power. FCE has been building molten carbonate fuel cells but it owns 40% or more of Versa Power that is developing solid oxide fuel cells.

  10. pure ignorance! a catalyst does not “break down” it is not changed chemically. IT IS A CATALYST (please look it up before writing drivel)! a catalyst becomes disorganized and can be reformed as its material is not lost. in fact they have an entire industry for making catalyst reformers. one fuel cell company offers to sell you a reformer as an upsell to your purchase.

Comments have been disabled for this post