Hydrogen and gas terminals: new plants, no operating history

New hydrogen pipelines have no operating experience

What happens. Hydrogen networks are being built or converted from gas pipelines. Operators know natural gas well, but hydrogen has a different density, speed of sound and leakage behaviour, and flows in the first years are low and uneven.

What it costs. Commissioning risk and cautious operation: pressures held higher than needed, compressors running in poor operating points, and long start-up phases while everyone learns the network.

Why the usual fix fails. Rules and tools from the natural-gas network are carried over and adjusted by experience. With no history of this gas in this pipe, experience cannot yet tell what a reasonable setting is.

How a fixed-price remote diagnosis checks it. From planned network data, early measurements and simulated load cases I check how pressure and flow behave in the network and which control concept holds them in range. The fixed-price remote diagnosis shows whether a network model with a control concept is worth building before the first large flows arrive.

Background: Computing unmeasured states for control

Hydrogen feed from electrolysers fluctuates with the wind

What happens. Electrolysers that follow wind and solar power deliver hydrogen in surges and lulls. The pipe or the customer behind them expects a steady supply at a defined quality.

What it costs. Pressure swings in the network, purity and moisture that vary with load, and electrolysers that run at poor efficiency or are switched off because the feed cannot be absorbed.

Why the usual fix fails. A buffer tank or a larger network smooths part of it, but it costs investment and still cannot be sized by rule of thumb. Simple limits on each electrolyser leave the units fighting each other.

How a fixed-price remote diagnosis checks it. From feed-in logs, pressure records and power forecasts I check how a coordinated control of several electrolysers and the buffer would have handled the swings. The fixed-price remote diagnosis answers whether coordinated feed control brings the grid quality you need and what it would ask of the existing control system.

Background: Economic predictive control

New ammonia import terminals have to be started up

What happens. Ammonia is to be shipped in as a hydrogen carrier. The terminals combine cold storage, pumps, heating and sometimes a cracker that splits ammonia back into hydrogen. Few teams have operated such a chain.

What it costs. Start-up phases that take longer than planned, losses from boil-off, and safety margins that keep the plant far from its efficient operating range.

Why the usual fix fails. Operating instructions are written from the design data and tested on first use. Each surprise is solved by a restart and a more cautious setting, which then stays.

How a fixed-price remote diagnosis checks it. From design documents, equipment data and any early operating logs I build a process model of the terminal and test start-up and part-load cases on it. The fixed-price remote diagnosis shows which control concept the terminal would need, and what to confirm before the first cargo arrives.

Background: Predictive control of process plants

Compressor and cooling energy at hydrogen refuelling stations

What happens. A refuelling station compresses hydrogen to very high pressure and pre-cools it before it goes into the vehicle. Compressors and chillers run on fixed rules, whether one car or a row of trucks is waiting.

What it costs. The cost per kilogram of hydrogen includes a large share of electricity for compression and cooling. Running the equipment in poor operating points or keeping it cold with no vehicle in sight adds to it, and so does wear on the compressors.

Why the usual fix fails. Larger storage tanks help, but are investment. Fixed thresholds for starting the compressor do not know when the next fill is coming or what power costs.

How a fixed-price remote diagnosis checks it. From station logs (fills, pressures, temperatures, compressor and chiller power) I check how a forecast-based operating strategy would have used storage, compressors and cooling. The fixed-price remote diagnosis answers whether optimised operation lowers the cost per kilogram at your station and what it would take.

Background: Economic predictive control

Hydrogen purity is checked only rarely

What happens. Hydrogen purity is verified by occasional samples or at the point of delivery. Between checks nobody sees whether impurities such as moisture or traces from the production process are creeping up.

What it costs. Fuel cells can be damaged by even small amounts of some impurities, and a damaged stack is expensive. Off-spec batches found late mean returns, disputes with customers and expensive analysis.

Why the usual fix fails. A continuous analyser for every impurity is costly and needs maintenance. More frequent sampling shortens the gap but never closes it.

How a fixed-price remote diagnosis checks it. From process data of production, drying and compression, plus the lab results you already have, I check whether purity can be estimated continuously from existing measurements. The fixed-price remote diagnosis shows which impurities could be tracked this way, and where a real analyser stays necessary.

Background: Computed measurements

Hydrogen metering must be proven accurate

What happens. Hydrogen is sold by energy content and mass, and the meters must suit hydrogen and a changing gas mixture. Operators have to show that quantities in the network and at the delivery point are measured accurately.

What it costs. Compliance work and uncertainty about which devices will be accepted. An inaccurate meter means a disputed invoice or a loss for one side of the transfer point.

Why the usual fix fails. Buying more expensive meters or putting two in series is a way out in the short term, but it does not prove how accurate the result really is under changing pressure, temperature and gas mixture.

How a fixed-price remote diagnosis checks it. From meter data, pressures, temperatures and gas compositions of your transfer points I check how large the measurement uncertainty is and whether a model of the line can back it up as proof. The fixed-price remote diagnosis tells you whether a model-based proof of accuracy is realistic for your metering stations.

Controls for liquefied natural gas distribution are hard to staff

What happens. Companies that distribute liquefied natural gas by truck, small terminal and bunkering station need control systems for cryogenic tanks, pumps, vaporisers and loading arms. Control engineers with cryogenic experience are rare.

What it costs. Projects wait for staff, and plants are started with simple controls that waste boil-off gas or run conservative loading sequences.

Why the usual fix fails. Adding another general automation engineer is not enough, because tank pressure, boil-off and loading interact in ways that a standard controller does not handle well. Training takes years.

How a fixed-price remote diagnosis checks it. From the plant drawings, tank and pump data and any operating logs I check how pressure, boil-off and loading interact and what a process and cryogenic control concept would need to look like. The fixed-price remote diagnosis gives you that view without a permanent hire.

Background: Predictive control of process plants

Liquefied natural gas terminals must send out more gas with limited flexibility

What happens. Terminals have become a larger part of the gas supply, and demand for send-out varies with season, storage level and the market. Tanks, vaporisers and boil-off gas compressors have to follow.

What it costs. Boil-off that is vented, flared or recompressed at a cost, and regasification energy spent at poor operating points. When the grid asks for more send-out, flexibility is limited by how the plant is run.

Why the usual fix fails. Fixed rules for the vaporisers and boil-off handling are tuned for a steady send-out. A change in demand is met by running extra equipment as a safety margin.

How a fixed-price remote diagnosis checks it. From tank levels, boil-off rates, send-out records and vaporiser data I check how a scheduling and control strategy would have handled a varying send-out. The fixed-price remote diagnosis answers whether boil-off management and send-out optimisation make sense for your terminal.

Background: Economic predictive control