Halfway Through 2026. Here Is Where the SyngaPure Project Actually Stands.

Summary

Zero-X's BMWK go-inno backed SyngaPure development has moved from slide-deck stage to a working engineering plan. System architecture is finalised across three integrated purification stages with an AI control layer. Component prototyping starts in Q3, beginning with the catalytic tar reforming stage — historically the failure point that sinks waste-to-chemicals projects. The two hard problems that remain are catalyst durability under variable feedstock and tuning the local LLM control loop.

Why Now

Syngas purification is not a luxury feature. It is the gate that most waste-to-chemicals projects die at.

The pattern is consistent across the industry. A gasifier produces raw syngas. The raw syngas contains tars, particulates, halogens, and sulphur compounds. Downstream processes — Fischer-Tropsch synthesis, methanation, catalytic conversion — require inlet gas specs measured in parts per million. The gap between what a gasifier outputs and what a chemical reactor requires is where most projects run out of engineering budget and investor patience.

Enerkem spent over 15 years and $200M+ solving this for their Edmonton plant. Fulcrum BioEnergy filed Chapter 11 in 2024 with the gas cleaning train cited as a contributing complexity. INERATEC and Velocys both run their own extensive gas conditioning systems. None of these solutions are packaged for distributed-scale gasification.

The SyngaPure project is Zero-X's answer to that gap. A three-stage purification train designed to match a single X-150 gasifier's output at 150 kg/h, packaged for containerised deployment. No custom building. No site-specific engineering. A catalytic tar reformer, a polishing stage for halogens and sulphur, and an AI control layer that adjusts operation as feedstock composition drifts.

What We Built

System architecture is finalised. Three integrated stages, with the AI control layer specification fully drafted. Partnership conversations with research institutions are progressing on catalyst and sorbent materials. Patent landscaping is complete. The roadmap is no longer a slide deck. It is a working plan with milestones, funding, and deliverables.

Component-level prototyping starts in Q3 2026. We are focusing first on the catalytic tar reforming stage, because that is historically the point where projects fail. Raw syngas from the X-150 carries 4-12 g/Nm3 of tars. Chemical synthesis processes require levels below 20 mg/Nm3. The reformer has to convert tars by a factor of 200-600x, continuously, over thousands of hours, on whatever feedstock the gasifier is fed that day. If we validate the tar reformer at industrial conditions, the rest of the system scales from there.

The COMETHA campaign already demonstrated 99.98% tar conversion using the X-150's existing process train over 1,939 hours of operation. The SyngaPure project takes that validation a step further — dedicated reformer design with catalyst and sorbent engineering optimised for long-duration, variable-feedstock operation.

This is BMWK go-inno backed development. German federal funding for applied industrial research, which means the project has passed a rigorous technical review and carries specific deliverables with defined acceptance criteria. This is not exploratory research. This is build-to-spec engineering.

Why It Matters

For the operator: three purification stages in a package that matches the X-150's containerised footprint. No custom engineering. No surprise integration costs at installation time. The gasifier and the purification train ship as one system.

For the investor: the tar reformer is the highest-risk, highest-value component in the syngas-to-chemicals chain. De-risking it at industrial scale unlocks Fischer-Tropsch, methanation, and SAF pathways for distributed waste gasification. The COMETHA campaign de-risked the gasifier. SyngaPure de-risks everything downstream.

For the engineer: the AI control layer is the novel piece — a local LLM adjusting reformer parameters as feedstock composition drifts, running on commodity hardware with decision latency measured in seconds. Tuning the balance between model size and response time is a genuine open problem. That is the kind of problem worth solving.

The Unglamorous Middle

The unglamorous middle of a deep-tech project is where most things break. It is also where the difference between a slide deck and a shipped product gets made. Catalyst durability under variable feedstock is the open question. Lab-scale results are encouraging, but industrial-scale validation will tell the real story. The local LLM control loop needs the right balance between model size and decision latency. Both are solvable. Both take time.

This is the work that doesn't make headlines but determines whether the project actually ships. More updates as we cross the next milestone.