What Will It Actually Cost? The Question That Started FuelRoute Pro

This is Part 1 of a series on building FuelRoute Pro — from a green methanol spreadsheet to an AI-integrated platform.

• Part 1 — The Spark (this article)
• Part 2 — Moving to the Cloud: Why We Left Squarespace and Vercel for Google Cloud
• Part 3 — The API: From Proof of Concept to Production Platform
• Part 4 — The Interface: Integrating with Claude and the Model Context Protocol
• Part 5 — What’s Next: The Future of FuelRoute Pro [Coming soon]

It started with a shipping supply chain and a number nobody could agree on.

At Thampico, we were working through the logistics of moving green methanol — estimating what it would actually cost to get the fuel from point A to point B. There were opinions. There were spreadsheets. What there wasn’t was a consistent, defensible starting point. Every analyst who touched the problem came back with different numbers, built on different assumptions, using different sources. The math was rarely wrong. The problem was that it was never the same math twice.

That frustration is where FuelRoute Pro began.

The problem hiding in plain sight

Energy transition runs on ambitious targets. It also runs, in practice, on logistics — and logistics cost money that is very hard to estimate before you’ve committed to a pathway.

Take a concrete example. A major West Coast port needed accurate cost figures to evaluate methanol as a bunker fuel for cruise ships. Sounds straightforward. In practice, the cost of a single round-trip methanol delivery by chemical tanker depends on the charter rate (which varies by vessel size and contract duration), the number of crew by position (a handy-size chemical tanker typically runs 22 people, from Captain down to Deckhands), tug and pilot fees at both ends of the route, fuel cost for the vessel itself, insurance, surcharges, and contingency for regulatory uncertainty. Our estimate for one such route came to over $635,000 per round trip before contingency.

That is a real number for a real route. And it is the kind of number that lives entirely inside the head of whoever built the spreadsheet last.

The problem we kept running into was structural. In the alternative fuels world, two teams modeling the same hydrogen transport corridor might come up with $3 per kilogram and $8 per kilogram, both using reasonable-sounding sources. Neither team is being careless. They’re each building from scratch, making independent calls on dozens of small assumptions: distance calculations, truck capacity, terminal fees, insurance conventions, carbon offset costs. The assumptions compound. The results diverge. Then the two teams spend a meeting arguing about whose spreadsheet is right rather than making a decision.

There are sophisticated tools in the energy sector — lifecycle emissions models, techno-economic assessments, long-range production cost projections. What didn’t exist was something that could answer the operational question: if I need to move 500 tonnes of ammonia from the Gulf Coast to the Midwest, what does that cost, and where does the money go?

Building the first version

The first version of FuelRoute Pro was a proof of concept built as a capstone project. It was designed to test one hypothesis: could you take a fuel type, an origin, a destination, and a transport mode and with tat information produce a cost breakdown grounded in something real?

Real road networks. Real routing algorithms. Market-informed pricing for the components that actually make up a delivered fuel cost: the commodity itself, transportation, terminal handling, loading and unloading, insurance, and carbon offsets.

We started with five fuels — hydrogen, methanol, ammonia, LNG, and diesel as a baseline — and two transport modes: truck and rail. The development process was methodical. We worked through logistics research and data compilation first, building out the cost framework before writing a line of production code. We ran case studies against real-world routes to validate that the output was directionally credible. When the formulas were wrong, we updated them before moving on.

The core insight that shaped every design decision was this: the energy transition doesn’t need another tool that claims to eliminate uncertainty. It needs a tool that makes uncertainty visible, structured, and discussible. The goal was never to deliver final answers. It was to give teams a credible place to start — fast enough to be useful in the conversation where the decision is actually being made.

What we discovered about how it gets used

When we put the proof of concept in front of people something happened that we hadn’t fully anticipated.

The comparative function turned out to matter as much as the absolute estimates. A single delivered cost figure is useful. Two delivered cost figures for the same route, calculated on the same basis, are something different — they’re the foundation of a real conversation. When everyone in the room is looking at the same model, the argument shifts. Instead of debating whose spreadsheet is right, you’re debating the assumptions. That’s a more productive argument. It moves the decision forward.

The moment this became clear: a team evaluating fuel options for an industrial facility was able to put truck versus rail, and LNG versus methanol, side by side in minutes. The output wasn’t definitive — it was consistent. Same routing logic. Same cost structure. Same underlying assumptions. For the first time, they could see not just what each option might cost, but why it cost that — which line items were driving the difference.

That’s the version of the tool that earned the feedback that turned a proof of concept into a product.

What comes next

Users wanted more fuel types, finer cost breakdowns, and reports they could share with stakeholders. They wanted an API to integrate into their own systems. They wanted the tool to work the way their decision-making process works — which is messier and more iterative than any single estimate can capture.

Building that meant rethinking the infrastructure, connecting live commodity price feeds, and eventually integrating AI in a way we hadn’t imagined when we started. This exercise also surfaced a set of hard technical problems — unit validation, pricing fallbacks, transport physics, data quality — that turned out to be more interesting than they looked.

That’s Part 2.

Next in the series: Moving to the Cloud: Why We Left Squarespace and Vercel for Google Cloud

FuelRoute Pro is built by Thampico LLC.