Ten Things to Know About the Bloom Energy Server

The Bloom Energy “Server” was officially announced today (and yes, I think its stupid to call it a server, its a generator). But the hype that has built in only three days (since Sunday’s 60 Minutes feature) is hard to sort out. So here are ten quick facts about the Bloom Energy Server.

1. Its not renewable energy. While it is possible to run the generator using biogas which would make it carbon neutral, there isn’t nearly enough biogas to provide the approximately 2,500 TWh per year of baseload (24/7) energy needs in the United States. So it will run off of natural gas (methane), which will produce CO2.

2. It can provide heat in addition to energy. The reaction that produces electricity also produces some heat (exothermic). The reaction also needs heat (1000 degrees C), but some heat can be bled off for combined heat and power (CHP) situations, supplementing traditional heating systems or pre-warming water before going into a hot water heater. How much heat is unknown, other than it does support CHP.

3. It emits slightly less CO2 compared to a modern natural gas power plant. A modern natural gas power plant will release approximately 822 lbs. of CO2 per MWh of generation (pg. 8). The Bloom box emits 773 lbs. Factoring in 10% (average) transmission losses between a typical power plant and your home, thats 904 lbs. The Bloom energy server reduces CO2 emissions by 14.5%. Its better, but its no miracle. And this doesn’t include any carbon output from issues related to distributing the natural gas city-wide, distributed repair work, or materials required for the Bloom Energy Server vs. one large natural gas pipeline and a large natural gas turbine power plant. To its credit, it is certainly cleaner than coal which releases about 2,800 lbs of CO2 per MWh on average in the US. If coal power is what it replaces, thats great.

4. You’ll still need the grid. Because it appears that Bloom is going towards baseload power (the turf of coal and nuclear), you’ll still need the grid to supply power when your electricity needs spike, like with air conditioning units or any electric appliances (range, stove, clothes dryer). This has implications however, as noted below in number 6. This need will likely be met by conventional natural gas turbine power plants or solar in appropriate climates.

5. The future price of $3,000 for a home unit is still comparatively expensive. While Bloom is correct in that you could recoup the costs within 3-5 years at that price point and current energy costs, things are not going to stay that way. Plus, automotive companies are working to create powerful fuel cells for vehicles. A fuel cell for a passenger vehicle from GM or Toyota (the two leading fuel cell contenders at the moment) would need to produce as much as Bloom’s current large unit but come at a cost between $5,000 and $7,000 to be economical to sell to the public; never mind the fact that Bloom’s unit is a large as the car itself, while the automotive solutions would be small enough to volumetrically replace a traditional gasoline engine. Even incorporating the steam reformation technology, and the technology to convert the DC energy to AC energy, the automakers are going to have a cheaper per kW solution than Bloom is discussing – currently $7,500/kW based on estimates and the $3,000/kW future target, by comparison natural gas turbine is $700-1000/kW capital cost (pg 17). Automakers are working on reducing precious metals required for their fuel cells, down to about twice the platinum required for a catalytic converter (10g Pt). Their plans are to introduce fuel cell vehicles in 2015, with broad (read: affordable) consumer options by 2025. Bloom’s only hope in this aspect is that their fuel cell lasts much longer than the fuel cells for automotive applications.

6. Your regular power bill will not go down as much as you think. Power companies have large fixed cost bases (transmission, distribution, power plants) and if the utilization of those items drops significantly, they will have to raise prices to cover the fixed cost base. They will essentially be spreading the same fixed costs over a smaller amount of kWh being generated and transmitted. Guess how your power company bills you? Per unit of energy you consume, so you would see cost per unit of energy go up. Alternately or in addition to that, they may raise their monthly service fee to remain connected to the grid.

7. When it breaks, you’re responsible to get it fixed! Suffice to say, there are some benefits to relying on the power company – when their power plants break they fix them. When yours breaks you have to fix it. You’ll still be able to rely on the grid but, as mentioned above, its likely you’ll want to get it fixed as soon as possible. Bloom has a 10 year warranty on their first units, but its currently unknown if it will reach its payback period within that 10 year period.

8. Increased natural gas prices could negate some of the savings. For those who rely on natural gas for heating and cooking, they could see their prices go up as demand increases. While the United States has ample natural gas reserves, that doesn’t mean they are easy to get to and cheap to access. Gas companies will be required to expend money to develop new resources, and they will be sure to pass the costs on to you. During the last boom cycle, natural gas prices went for as high as $9/MMBTU which would translate to 6c/kWh for fuel only (not including capital or maintenance costs).

9. A smart grid will be needed to regulate your personal Bloom Energy Server(s). The smart grid, which is slowly developing around the world, will be needed to regulate the output of the personal Bloom Energy Servers. If not, they will at the very least need to be able to sense the load of your house and not put energy back onto the grid. Some electrical equipment in the grid (at the substation level) cant handle energy flowing backwards, if all the Bloom boxes put energy back onto the grid at night, it could harm the distribution infrastructure, to say nothing of individuals who erroneously think it would be a good idea to have multiple Bloom Servers so their peak use would be covered but provide a large surplus back to the grid.

10. They don’t have the right business model. This probably isn’t that big of an issue now, but as Bloom grows, they aren’t going to be able to monitor the power generation boxes in every home and business in America. This is a job for the utilities. Bloom will need to adjust its business model to join with utilities or third parties so they can monitor the devices and provide services like maintenance and repair, as well as utilities controlling how the boxes get installed and attached to their grid.

So thats it. I hope a healthy bit of skepticism douses the hype this new breakthrough got. It looks to be an important step in the right direction. There are certainly niche cases where something like this would be incredibly valuable. But as far as having one in every home in the country generating the baseload energy, that seems a bit far fetched, at least in the next 10-15 years and costs retreat dramatically.