Comment by Hypx | Hacker News Reader

Comment by Hypx | Hacker News Reader

Parent Full thread

1) Main strategy to avoid this is to just coat pipelines with plastic or make them out of plastic. Not all metals are equally susceptible to this process, and newer alloys are being developed that can strongly resist this phenomenon. While it is a problem, it is a solvable problem.

2) What I’ve read is that they’ll build hydrogen sensors and alarms for this.

3) This is hugely overstated for two reasons. One, just because an process contains inefficiencies, doesn’t mean the entire process is invalid. After all, gasoline powered cars throw away more than 80% of the original energy found in crude oil as waste heat. Fuel cell cars may waste more energy than battery cars, but it is less than ICEVs. Also, you do want to have excess heat sometimes.

Second, most of the sources of inefficient are solvable and/or were wildly exaggerated to begin will. Many arguments against hydrogen seem to cite some 20-30 year old source, and had some extremely pessimistic projections like hydrogen tanks leaking 20% or more of their fuel. Most of these projections haven’t been borne out.

4) Why do electrolysis plants need to run at 100% utilization rates? In a renewable world, almost nothing will have 100% utilization. This is not really a big deal since electrolyzers can be turned on and off quickly.

5) That’s true of any subsidy. One of the biggest complaints of the subsidy regime is how wasteful it is, and how poorly optimized it is for reducing GHG emissions. It will be almost a relief if hydrogen gets of those subsidies, since right now it is an important necessary technology that’s getting close to zero subsidies.

6) We don’t really need two overlapping infrastructures. Limit each infrastructure to where it is most effectively used.

Ad 1) But is that realistic, given the existing LNG infrastructure? The story being told is that we can just pump H_2 into the existing LNG pipes.

Ad 2) Yes, in processing environments. The story that is told to us is that we can replace LNG with H_2 to heat houses (with minimal investments).

Ad 3) True, ICEs are grossly inefficient. However, as I've seen it, EVs are still twice (?, reference needed) as efficient as H2. Especially for renewable, efficiency is key, as the whole production line competes with the price of oil.

4) Capital Expenditures. The YoY ROI is obviously very strongly influenced by the utility rate.

5) Green hydrogen as it stands is heavily subsidized and it needs to be. The capital expenditures for a hydrogen filling station is more than 50% covered by subsidy. And that is optimistic. So, close to zero subsidies, is not what I am seeing.

Ad 6) This might be true for industry usage, but not for consumer usage.

1) You can blend hydrogen with natural gas. To go 100% H2, you may need new pipes, but this isn't an insurmountable problem.

2) These sensors aren't that big. It'll probably be like fire alarms but for hydrogen if we go that route.

3) The gap is shrinking over time, as fuel cells and electrolyzers get better. Batteries are basically maxed out on efficient so they won't improve by much. Loosely spoken, the gap should shrink to about a 50% advantage for the EV in the long run. That means EVs may always make sense for short ranged trips, but for anything that needs big batteries the cost of the batteries probably will outweigh electricity savings.

4) If electrolyzers get really cheap this may not hold. I suspect that optimizing to 60-70% utilization is more than good enough.

5) Relative to the competition, it is tiny. We're subsidizing wind, photovoltaics, batteries, etc., to the tune of tens of billions, but for hydrogen it was in the hundreds of millions. Though this is starting to change as well speak as hydrogen is starting to get bigger subsidies.

>Why do electrolysis plants need to run at 100% utilization rates? In a renewable world, almost nothing will have 100% utilization. This is not really a big deal since electrolyzers can be turned on and off quickly.

The problem is capex. I am not at all an expert here, but from what I have read electrolyzers are at present very expensive, so to pay off their capex they need to be run continuously. Perhaps that will change in the future.

I'd imagine it should change in the future. We're looking at something like a 10,000x increase in production so current models won't hold in the future.

Reply on news.ycombinator.com