Would Cheap, Clean Hydrogen Make Fuel Cells Superior to Batteries?

[mheslep]

Wow! You sure ask a lot of question... are you trying to set up a hydrogen refueling station?

I don't mean to burden you individually with Q&A sessions. You did suggest the H2 economics were such in post 7 that they could tip over the material handling business, a substantial claim given a existing $2-4 billion business using batteries or petroleum power. But maybe so. I've been through the National Academy reports on H2, etc but they admit to many estimates that are difficult to check. Here we have a case we can check. How? By modelling with as much detail as possible and crunching the numbers. We're engineers, that's what we do.

On the flik, you ponted out all the bits and pieces. Tube trailer, compressor, dispencer. The one shown in the flik is an older style with electric hook ups. (forgot to mention that) The new ones have an infrared connection on the dispencer nozzle so you don't have to hook those wires up anymore.

Yea, the tube trailer is going to cost a LOT more for a kg of hydrogen. As mentioned before, it's about 3 to 4 times the cost, primarily because of all that extra metal that has to be dragged around. ...

That makes sense, yet the WholeFoods warehouse went with compressed gas and not liquid storage. Thus I assume they found the on site liquid storage costs to be prohibitive, even given the cheaper transportation via liquid H2 trucks.

 

[mheslep]

Here's a new pricing data point on the fuel cell end. The well funded startup Bloom Energy finally went public with their product - a 100kW high temperature ceramic fuel cell. No noble metals, no reformer, runs on various hydrocarbon gasses, 50% efficient. Fuel cell price is ~$7500/kW (before subsidies), they expect $3000/kW. At least for the moment they are far short of the $100/kW for an ICE or an electric motor.

http://www.luxresearchinc.com/blog/2010/02/is-bloom-energy-a-better-place-redux/
Wired and 60 minutes also did pieces on them yesterday and Sunday.

 

[Topher925]

Here's a new pricing data point on the fuel cell end. The well funded startup Bloom Energy finally went public with their product - a 100kW high temperature ceramic fuel cell. No noble metals, no reformer, runs on various hydrocarbon gasses, 50% efficient. Fuel cell price is ~$7500/kW (before subsidies), they expect $3000/kW. At least for the moment they are far short of the $100/kW for an ICE or an electric motor.

Just an FYI, all modern SOFCs contain no noble metals. Bloom's is no different than anything else out there other than its very inefficient (most SOFCs are ~65-70% efficient).

On another note, there is no way to realistically compare a SOFC to an ICE designed for a car. Two very different technologies, for two very different applications. Those prices also don't include the cost savings of fuel over the life of the system since SOFCs require about 3 times less fuel than your ordinary ICE and about 4 times as less than your a turbine engine. SOFCs also typically require significantly less maintenance.

And $3k/kW is outrageously expensive after including subsidies. You can go out and buy a PEM cell with the same efficiency for $3k/kW with no subsidies what so ever and even that's on the expensive side.

http://www.horizonfuelcell.com/store/h5000.htm

 

[mgb_phys]

On another note, there is no way to realistically compare a SOFC to an ICE designed for a car.

At current efficencies if you ultimately want rotary motion you are probably better off with an ICE than a FC+Electric motor.

 

[mheslep]

Bloom's is no different than anything else out there other than its very inefficient (most SOFCs are ~65-70% efficient).

We're talking about an in production FC here, not a lab prototype. Bloom makes numerous claims other than efficiency about its fuel cell. Are saying you have reviewed all the salient claims, compared them to actual SOFC products available and found them no different? Who has an SOFC product on the market at 65%?

On another note, there is no way to realistically compare a SOFC to an ICE designed for a car. Two very different technologies, for two very different applications. Those prices also don't include the cost savings of fuel over the life of the system since SOFCs require about 3 times less fuel than your ordinary ICE and about 4 times as less than your a turbine engine.

I didn't claim lifecycle costs, I stated the upfront kW cost only, as reported in the reference. How do you get FC eff * E-motor eff (~92%) = 3 X 25% ICE eff?

SOFCs also typically require significantly less maintenance.

Than what? A diesel backup/load levelling generator?

And $3k/kW is outrageously expensive after including subsidies. You can go out and buy a PEM cell with the same efficiency for $3k/kW with no subsidies what so ever and even that's on the expensive side.

http://www.horizonfuelcell.com/store/h5000.htm

Yes $3k/kw is expensive, but that PEM is not comparable. No reformer to run off natural gas. No three phase high power AC inverter. No packaging for outdoor operation. Ten percent less efficient.

 

[Topher925]

At current efficencies if you ultimately want rotary motion you are probably better off with an ICE than a FC+Electric motor.

How do you figure? FC coupled with traction motors are significantly more efficient than any ICE engine available.

We're talking about an in production FC here, not a lab prototype.

No, I'm talking on the market commercialized SOFCs. SOFCs are very popular in applications where reliable steady state power is required. Heres an example of some manufacturers but many more can be found with a simple search.
http://www.fuelcellmarkets.com/fuel....html?q=Systems,Solid_Oxide_Fuel_Cell_Systems

I didn't claim lifecycle costs, I stated the upfront kW cost only, as reported in the reference. How do you get FC eff * E-motor eff (~92%) = 3 X 25% ICE eff?

SOFCs like the ones Bloom has made can only be used for stationary applications and APUs. In these applications you go from fuel to electricity in with a FC. ICEs require you go from fuel to generator to electricity. However, if you do consider a transportation application, SOFC+Motor:65% x 95% >> ICE:25%.

Than what? A diesel backup/load levelling generator?

Yes.

 

[mheslep]

No, I'm talking on the market commercialized SOFCs. SOFCs are very popular in applications where reliable steady state power is required. Heres an example of some manufacturers but many more can be found with a simple search.
http://www.fuelcellmarkets.com/fuel....html?q=Systems,Solid_Oxide_Fuel_Cell_Systems

Yes I'm aware there are SOFC's companies, though it is not clear any of them are are past pilot projects and into production. But that was not my question. I was asking for a reference to your claim that most "SOFCs are ~65-70% efficient" - products not research prototypes - , and that "Bloom's [SOFC] is no different than anything else out there." I can't find anything that exceeds 60%, and even that is only at a specific power operating point less than full load. BTW, Bloom's exact spec is ">50%"

Yes.

Having designed one (diesel electric APU) into a commercial building, and having reviewed the FC reliability comments from the National Academy report, I doubt that is true. More reliable than gasoline-generators maybe, not diesel APUs. I expect the reliability comparison is a wash.

 

[russ_watters]

Those prices also don't include the cost savings of fuel over the life of the system since SOFCs require about 3 times less fuel than your ordinary ICE and about 4 times as less than your a turbine engine. SOFCs also typically require significantly less maintenance.

Where do you get the cost savings over a turbine? A commercial gas turbine is around 30% efficient, so a 60% efficient fuel cell would use half, not 1/4 the fuel.

In the US, for commercial use, a GTE would almost certainly include heat recovery to provide all hot water and heating in addition to electricity, so the all-up efficiency may end up being better than a fuel cell.

How does a fuel cell perform at part load?

GTEs are somewhat of a niche market right now, but one very good niche is sewage treatment plants, which use megawatts of electricity (to run massive pumps) and tons of heat (to keep their vats of crap warm) while producing lots of methane. Traditionally, they'd take grid power to run the pumps and capture some of the methane for heat (maybe 25-50%), flaring (burning off) the rest. A GTE can provide both the electricity and heating.

 

[Topher925]

can't find anything that exceeds 60%, and even that is only at a specific power operating point less than full load.

I believe that 60% is total power conversion, including transformers and/or converter. I have lots of info about SOFCs that can operate at those efficiencies, unfortunately none of it I can share on the internet. I'll try to find some other sources later.

You may be right about diesel and SOFC reliability being a wash, I really have no info about diesel generators but for SOFCs the only components that ever really need maintenance are pumps.

Where do you get the cost savings over a turbine?

http://www.netl.doe.gov/publications/proceedings/03/seca/daviddaggett.pdf (slide 15)

In the US, for commercial use, a GTE would almost certainly include heat recovery to provide all hot water and heating in addition to electricity, so the all-up efficiency may end up being better than a fuel cell.

Not going to happen. SOFCs (and all high temp FCs) produce high quality heat just like turbines except they produce a little more of it and are often used for co-generation as well.

http://www.greencarcongress.com/2009/12/tmc-sofc-20091218.html
http://www.energy.siemens.com/hq/en/power-generation/fuel-cells/sofc-commercialization.htm

Combustion of gases in turbines causes dissociation of products reducing temperature and the amount of heat that could have been used for co-generation. This is something that can be avoided with SOFCs.

How does a fuel cell perform at part load?

Depends on the design, but usually just fine. Although FC efficiency is directly proportional with overpotential. SOFCs usually don't care at what load they operate at as long as its enough to stay warm. What does matter however, is the rate at which that load changes. Like almost all energy conversion and storage devices, SOFCs like to operate at a steady state constant load. If the load operates with large transients, you've got problems.

A GTE can provide both the electricity and heating.

See above.

 

[russ_watters]

http://www.netl.doe.gov/publications/proceedings/03/seca/daviddaggett.pdf (slide 15)

Did you see slide 14? Your data point is for a case where the jet APU operates far outside its design range, but doesn't say exactly what that range is. It is unreasonable to assume 15% efficiency from a GTE would be typical. Slide 14 shows that in normal operating conditions, the difference is only 45%.

Not going to happen. SOFCs (and all high temp FCs) produce high quality heat just like turbines except they produce a little more of it and are often used for co-generation as well.

If they do heat recovery, fine, but that means the all-up efficiency would end up being about the same.

But you do have that first part backwards: since fuel cells are more efficient, the waste heat is lower than in a GTE (generally a good thing, I'm just pointing it out for clarity).

http://www.greencarcongress.com/2009/12/tmc-sofc-20091218.html
http://www.energy.siemens.com/hq/en/power-generation/fuel-cells/sofc-commercialization.htm

Your second link claims an all-up efficiency of 80%, which is exactly the same as a GTE operating at the lower heating value*. http://www.capstoneturbine.com/prodsol/solutions/chp.asp

Since the chemistry is exactly the same, this isn't surprising that they both hit the exact LHV efficiency that is typically possible in any fuel fired heating application (such as your home boiler or furnace). Not sure why they can't condense the water out and go for the HHV, though...

Combustion of gases in turbines causes dissociation of products reducing temperature and the amount of heat that could have been used for co-generation. This is something that can be avoided with SOFCs.

I'm not following. The chemistry of these things is is exactly the same:

[CH4 + 2 O2 => 2 H2O + CO2 + E]

They do the same thing: they both burn methane in air.

*Edit: Actually, in the literature I linked, they are switching back and forth between the HHV and LHV, which is a little misleading: what they are really claiming is 80% of the LHV efficiency, or 75% of the HHV. Your first link shows both for a fuel cell: 85% HHV, 77% LHV.

 

[mheslep]

How does a fuel cell perform at part load?

This Australian SOFC model reaches peak efficiency at about 80% rated load, falls off slightly at 100%, and falls off drastically at below ~40% load.

http://www.fuelcellmarkets.com/images/articles/443/graph.gif

It's not clear it this company is actually in production mode yet.

 

[russ_watters]

Thanks. Put another way, at 25% ouptut, the efficiency drops by 40%. For the turbine I linked above, the efficiency drops by 30% at 25% load. The curves are not the same shape, though, and it looks like their curves cross somewhere around 10% load (ie, below 10%, the fuel cell isn't losing as much of its peak efficiency).

 

[mheslep]

..
Your second link claims an all-up efficiency of 80%, which is exactly the same as a GTE operating at the lower heating value*. http://www.capstoneturbine.com/prodsol/solutions/chp.asp

Also, one can actually buy a micro CHP from Capstone and obviously larger scale co-generation has long been available. That Siemen's SOFC on the other hand is not in production.

The chemistry of these things is is exactly the same:

[CH4 + 2 O2 => 2 H2O + CO2 + E]

They do the same thing: they both burn methane in air.

Not quite. Don't forget the combustion side reactions due to high temperature in the presence of Oxygen which, in addition to their resultant pollution, consume energy. E.g.
O2 + 2 C + E => 2 CO
Same for NOx, etc.

 

[Topher925]

Did you see slide 14?

Yes I did. I didn't include it because slide 14 refers to electrical power generation while the engine is driving the plane. In other words, the engine isn't only working to produce electricity, its producing thrust and electricity. This will give a value of apparent efficiency making it not comparable to electrical or co-generation only power generation.

But you do have that first part backwards: since fuel cells are more efficient, the waste heat is lower than in a GTE

I think I got it forwards, this is a good characteristic to have. In most applications its always more favorable to produce more electricity than heat since electricity is a very versatile form of power and heat (especially at lower temperatures) is not.

The chemistry of these things is is exactly the same:

[CH4 + 2 O2 => 2 H2O + CO2 + E]

They do the same thing: they both burn methane in air.

The chemistry is most definitely not the same. SOFCs don't "burn" anything. What you posted is the IDEAL global reaction and not what you would find in a real application.

For a SOFC, a more realistic equation would look something like this,

CH4 + 2 (O2 + 3.72 N2) => 2H2O + CO2 + 7.52 N2

For a GTE, especially a higher efficiency one (higher temp), it would look something more like this,

CH4 + a (O2 + 3.75 N2) => b H20 + c CO2 + d N2 + d CO + e NO + f NO2 + g H2 + etc.

http://en.wikipedia.org/wiki/Dissociation_(chemistry)

 

[mheslep]

http://en.wikipedia.org/wiki/Dissociation_(chemistry)

As I recall and the Wiki indicates Dissociation refers to ionic bonds, not the covalent bonds broken and made in combustion.

 

[russ_watters]

Not quite. Don't forget the combustion side reactions due to high temperature in the presence of Oxygen which, in addition to their resultant pollution, consume energy. E.g.
O2 + 2 C + E => 2 CO
Same for NOx, etc.

[and for Topher]

Granted, but just how much are we talking about? According to GE, all of those are in the PPM range, with a total of all of them of less than 1 part in 1000, so the efficiency hit you take is several significant digits away from the decimal point in the efficiency calculation. It isn't enough to be relevant.

 

[russ_watters]

Yes I did. I didn't include it because slide 14 refers to electrical power generation while the engine is driving the plane. In other words, the engine isn't only working to produce electricity, its producing thrust and electricity. This will give a value of apparent efficiency making it not comparable to electrical or co-generation only power generation.

The efficiency number does not include the propulsion of the plane. It says that explicitly on the slide: ""40-45% Jet-A to electrical efficiency"

If that bothers you, just drop this special case and compare the fuel cell to a similar product such the capstone turbine I linked. 15% is not a fair/realistic efficiency for a typical commercial gas turbine generator.

I think I got it forwards, this is a good characteristic to have. In most applications its always more favorable to produce more electricity than heat since electricity is a very versatile form of power and heat (especially at lower temperatures) is not.

No, I was referring specifically to your claim that more waste heat is available with a fuel cell. This is simply wrong.

The chemistry is most definitely not the same.

From the post above: when the difference in the chemistry is on the order of hudredths or thousandths of a percent, it isn't useful to call it different (at least when efficiency is what you are comparing).

As far as what to call what a fuel cell does, different sources say different things. Some sources say any reaction that involves oxidation and releases energy (such as rust) is a combustion reaction. Some say it must happen fast and/or involve heat and light as the energy released. Whatever. I'm less concerned with the label than I am about the similarity between the chemistries.

 

[mheslep]

[and for Topher]

Granted, but just how much are we talking about? According to GE, all of those are in the PPM range, with a total of all of them of less than 1 part in 1000, so the efficiency hit you take is several significant digits away from the decimal point in the efficiency calculation. It isn't enough to be relevant.

Yep, that makes sense, should be small.

 

[Topher925]

If that bothers you, just drop this special case and compare the fuel cell to a similar product such the capstone turbine I linked.

Ok.

No, I was referring specifically to your claim that more waste heat is available with a fuel cell. This is simply wrong.

Sorry, I should have been more clear. What I meant was that the amount of energy that can be utilized per unit of fuel is greater. This is ultimately because you are not forming products which are generated by an endothermic reaction removing usable energy nor do you have to deal with them after your process.

From the post above: when the difference in the chemistry is on the order of hudredths or thousandths of a percent, it isn't useful to call it different (at least when efficiency is what you are comparing).

I'll agree that the amount of heat lost isn't significant but in many cases dissociation can drop an ambient flame temperature by a hundred degrees or so. Granted this isn't a big deal for turbines as flame temperatures are usually as high as alloys can take but the pollutants should be considered. After all, things such as catalytic converters on cars do reduce engine efficiency (and increase cost).

As far as what to call what a fuel cell does, different sources say different things.

Fair enough. I'm not concerned about semantics.

 

[russ_watters]

Sorry, I should have been more clear. What I meant was that the amount of energy that can be utilized per unit of fuel is greater.

Ok, fine.

This is ultimately because you are not forming products which are generated by an endothermic reaction removing usable energy nor do you have to deal with them after your process.

No, as I pointed out, the differences in the chemical reaction are less than one part in a thousand. Essentially all of the difference in efficiency is due to the thermodynamics of a brayton cycle heat engine vs direct electrical conversion of a fuel cell.