Don’t Be the Guinea Pig
Last week I wrote about ITM power’s electrolyzer and how its much lower Total Installed Cost will result in lower H2 prices in any reasonable commercial setting. The major caveat was that reliability is paramount – sacrificing reliability to achieve lower cost is a non-starter in most industries. The end result is that hydrogen will cost more – because backup hydrogen deliveries are $20-$30/kg if they are even available. This quickly escalates costs and will either destroy margins or alienate the customer if you pass it on.
Gray hydrogen should cost $1/kg to produce, $2/kg to deliver, and $4/kg or less to fuel, assuming high utilization of each of the assets. Most of the equipment is trash on every part of this value chain – meaning there are single points of failure everywhere, and utilization is terrible. Let’s say that a fueling station is out 50% of the time – that means that the equipment to produce and distribute hydrogen is used half the time even if they are in operable condition and the CapEx contribution to cost doubles for every part of the value chain. With three single points of failure, every part of the system could double three times, for an 8x increase in cost. Our $7/kg H2 has suddenly become $56/kg.
While the 50% unlinked downtime across all parts of the value chain is exaggerated, the outcome is not. Delivered hydrogen is $20-$30/kg in much of the US. There are many contributions to this high cost, but reliability of equipment is one of the major ones.
These are just a few of the issues with reliability. Others include:
1. Complete project failure. Every end use requires reliable hydrogen. Most project failures are from hydrogen not being delivered. An $800,000 bus that can’t run because there is no hydrogen is a major waste of money. If there is hydrogen but the equipment doesn’t work, the problem remains
a. This goes for every single use case. Trucks, cars, power, agriculture. Failed delivery or ability to use the hydrogen will lead to failure
2. Long-term bad reputation: The cause of failure doesn’t matter to consumers. Every failure is another notch that H2 detractors track. They don’t care about successes, they don’t care about reasons for failures, a failure provides all detractors with examples of H2 failure
3. Reduction in investment and development: Every failure increases the perceived risk of hydrogen. The cost of this is higher ROI required by developers and investors – meaning most projects won’t get built and the ecosystem continues to shrivel.
Being the guinea pig for a company that has untested equipment will not only cause you major fiscal losses, it will also set back the entire industry. It continues to enable bad actors in the space.
Hallmarks of high risk vendors
• A startup that does lobbying before they have a product – do not go near this
• Companies with more marketing than engineering– their marketing team probably is selling hardware that the engineers are very uncomfortable with selling because it doesn’t work yet
• Hardware without extensive testing that mimics real-world conditions – IE start-stop cycles, temperature swings, input variability like a few ppm extra water in the H2 etc
What is your best solution
Hire an experienced advisor – it will cost $10-$30k and save you the entire cost of a multi-million dollar project. Hiring an experienced team might double or triple the cost, but the savings are worth it.
The major issues, however, is knowing who has a real understanding of what works and what doesn’t. If your advisor doesn’t have hands-on experience in developing tech from lab scale to commercialization, you’re missing something. Procurement folks excel in established industries and fail in nascent industries. It usually takes a product strategy person with some mix of lab, machine shop, or factory experience to truly probe reliable systems. These are rare. A person who can think both in engineering terms and commercial terms is usually an entrepreneur building their own companies – but some of us like the more relaxed lifestyle of independent consulting.
Find startup-experienced procurement - Another great option is to find a procurement person in nascent industries that worked on a team that successfully deployed several projects. I’m not talking about VCs here, you need someone at a corporation that specifically helps develop new tech.
With the downturn in H2 after 45V was cancelled and hubs are under attack, you will see a lot of people without the right experience doing this sort of consulting. If they were a paper pusher in their old job, they are still paper pushers and should not be your first choice for one-off engineering consulting.
Work with a company that has a balance sheet or sufficient funding to eat some of the testing cost – A startup with new hardware that isn’t well tested should be eating a significant part of the cost of deployment – and it should not be on a critical project for you. Look for a startup that is open about taking a serious haircut for their first deployments
Other Solutions
1. Find a “full-wrap” EPC that will guarantee production – these don’t exist in H2 yet or the cost is 3x the hardware cost
2. Purchase from a company that guarantees their product and has an excellent balance sheet. Note that startups and companies on the verge of bankruptcy explicitly don’t count – you are their lifeline and if their hardware doesn’t work
3. Procure hardware that has insurance built in – which you will likely see more of in the coming years as insurers build out teams to assess reliability so you don’t have to
Other solutions I recommend against
1. Industrial gas companies. They do nine figure projects. If you aren’t dealing with 10+ tons per day, you don’t matter. They have demonstrated that they will not honor their contracts, claim force majeure for issues they create, and then threaten enough legal defense that it isn’t worth going after them
2. Strategy consultants – don’t spend hundreds of thousands of dollars to hire an agency that doesn’t have actual boots-on-the-ground experience. There are good consultancies for H2 out there, but from my direct experience McKinsey and BCG are not among them
3. EPCs – EPCs are great once they have moved up the learning curve. No experienced people work at an EPC when they can get better salary and stock options at just about anywhere. EPCs are amazing at things they have done many times before. They are terrible at new tech. I have seen Fluor Corp give high ratings to hardware based on only a process flow diagram that is commercially unfeasible and questionably viable
When to not follow these rules – if you’re an experienced multinational corporation with extensive experience in uplifting companies
Shell had many attempts at making H2 work. They took a shotgun approach and made many partnerships and investments before the new CEO instituted fiscal responsibility.
ITM was one such partnership. Their original 10MW of electrolyzer at the Rheinland refinery had a lot of learning to go through. That project was a tech test – and Shell knew it could result in issues. Five years later, Shell chose ITM to expand the same facility to 100MW after years of working with the company to enhance the equipment reliability and serviceability.
Most companies don’t have the balance sheet, patience, and experience to build up an external team and technology. Shell is one of the few that has the skill to develop hardware and teams, but also recognizes that they cost 10x-50x more than a small team. In other words, the Shells, Exxons, and Equinors of the world can level up a vendor. Most companies and investors can’t.
This is a problem in every ecosystem
Wework lost ~$12B billion dollars. Elon has been promising autonomous vehicles coming “next year” for the last decade. Fusion has been 20 years away for 50 years. Elizabeth Holmes at Theranos reached an $9B valuation and had an advisory board that looked more likely to take over a small country than to guide hardware development (George Shultz, former US secretary of state; James Mattis – former US general; Henry Kissinger – that former US secretary of state; Gary Roughead – former US navy admiral, two senators, etc).
Don’t feed the trolls
Industrial hydrogen has been highly successful for almost 100 years now and is critical to our food supply and chemicals, demonstrating with proper product-market fit H2 is world-changing. Retail hydrogen can work, as can hydrogen in new industrial use cases. The key is product-market fit, and most of the hydrogen companies you have heard of nail the marketing while neglecting the core product.
Industrial H2 literally runs our agriculture and supports our entire chemicals industry. Commercial and retail H2 will have its day. Just don’t be the guinea pig for companies with unproven tech – you’ll go bankrupt and also perpetuate the failures that hold retail H2 in a quaqmire.
Real world examples of what it means to be a guinea pig – AKA “Story time with grandpa”
The main point of the article is done – only keep reading if you want some lore from recent H2 issues
Nel leads California’s failure to develop hydrogen refueling
Before I joined Shell as the head of hydrogen product strategy, the bizdev and procurement team had chosen Nel Hydrogen for seven hydrogen fueling station. Nel, a company that is more marketing than engineering, made false claims about their hydrogen compressor reliability – mostly because the system had never been deployed before and had not faced realistic reliability testing. Nel used both Shell and Iwatani as guinea pigs to test their hardware.
Despite having redundant compressors, Nel’s station hardware uptime was about 70%. Owing in part to this, Shell later abandoned retail hydrogen in the US, Nel is being sued by Iwatani for false statements.
Acquiring a compressor with an insurance plan that it 50% of CapEx
For a project to refuel heavy duty trucks, I came in late in the procurement process to do due diligence. I saw that the $500k hydrogen compressor had a $250k/yr maintenance policy if the buyer wanted the equipment to have 95% guaranteed uptime. The clear picture is that the equipment will be constantly breaking and they need a full-time person just to maintain it.
If the annual maintenance for 95% or higher uptime is more than 10% of the hardware cost – this hardware is still under development.
Not learning from others
I was on a call with a potential client and recognized the compressor they had purchased for their pilot. I told them their compressor would fail. They didn’t want to hear it. They weren’t paying me, so I wasn’t going to push it.
Several years later the compressor still didn’t work and the company failed. The investors took over and sacked all the leadership. Their hubris cost $40M in investment. A $15k contract would have saved that investment.
Acquiring hardware from a company that has no balance sheet – hardware failure means you cannot get equipment serviced
Shell acquired a small hydrogen production unit 20 years ago for a fueling station. The company went bankrupt. The developer of the hardware then charged egregious hourly rates to consult with Shell on how to keep the hardware operating – with the original company going bankrupt the engineers were free to charge whatever they wanted.
This is a risk that every company runs if they work with a small company – if the company fails then there is no one to service the equipment. It becomes a stranded asset. Due diligence is extremely important here. A $15k to $30k outside due diligence can save you this hassle.
