Chemical plants: late lab values, loops in manual, wide margins

Many control loops run in manual, detuned or oscillating

What happens. On the operator screens, a good share of the loops are in manual, or in automatic with settings so soft they barely act. Others swing around the setpoint all shift. Operators know which ones to watch and nudge them by hand.

What it costs. Swinging temperatures and flows push product towards the specification limit, so targets carry extra margin. Valves that never rest wear out early, and steam and cooling water are spent fighting the swings.

Why the usual fix fails. Retuning by trial and error on the running plant is slow and risky, so it stops after the worst loop. Sticking valves, noisy measurements and interacting loops mean settings found one day fail when throughput or feed changes.

How a fixed-price remote diagnosis checks it. From historian data (setpoints, measurements, controller outputs, modes) I rank the loops by how much they cost you and look for the cause of each: a sticking valve, a poor measurement, interaction with another loop, or simply wrong settings. The fixed-price remote diagnosis tells you which loops a model-based retuning would fix and which need mechanical work first.

Product quality depends on who is on shift

What happens. The same unit makes better product on one shift than on another. Experienced operators move setpoints early because they feel what is coming; others wait for the lab value and then correct hard. Shift handovers are where the trends jump.

What it costs. Variation between shifts means wider margins for everyone, more off-spec material to blend or rework, and complaints that are hard to trace. When the best operators are on holiday or leave, quality drops with them.

Why the usual fix fails. Written operating instructions capture the routine, not the judgement. Training helps, but the knowledge is in how a person reads several trends at once, and that does not fit in a manual. Locking setpoints just moves the variation into the product.

How a fixed-price remote diagnosis checks it. I compare the historian data of good and poor shifts: which setpoints were moved, when, and in response to what. The fixed-price remote diagnosis shows whether a process model can turn what the best shifts do into recommended setpoints for every shift, and which measurements that would need.

Maintenance by calendar because nobody knows the real wear

What happens. Pumps, heat exchangers, agitators and valves are serviced on fixed intervals, because nobody can see how worn they really are. Some come out in perfect condition; others fail between two inspections. There are fewer technicians every year to do the rounds.

What it costs. Unplanned stops when something fails early, shutdown time spent on equipment that did not need it, and spare parts bought just in case. Each surprise failure pulls the few remaining technicians away from planned work.

Why the usual fix fails. Extra vibration and temperature sensors on every machine cost money, need wiring and produce data that someone has to watch. Shorter intervals add work the team has no staff for.

How a fixed-price remote diagnosis checks it. Much of the wear already shows in signals you log: motor current against flow, pressure drop against throughput, valve position against flow. I look at those histories next to your maintenance records. The fixed-price remote diagnosis answers whether a condition value computed from existing signals would have seen past failures coming, and for which equipment.

Lower emission limits and more monitoring, with analysers that can fail

What happens. German air-quality rules and the revised European industrial emissions rules bring lower limits and more continuous monitoring. The plant now depends on analysers to show it stays below the limit. When an analyser drifts or fails, operators reduce load or open a margin until it is back.

What it costs. Analysers are expensive to buy and maintain, and an outage can cost production. Running well below the limit to be safe costs fuel, reagent or throughput every hour.

Why the usual fix fails. A second analyser doubles the cost and still fails in the same conditions. A bigger margin to the limit is the usual answer, and it is paid for every day.

How a fixed-price remote diagnosis checks it. From your analyser records, fuel and load data, temperatures and reagent flows I check how well emissions can be computed from what the plant already measures. The fixed-price remote diagnosis tells you whether a computed value can bridge analyser outages and support running closer to the limit, and what the authority would need to see.

Background: Computed measurements

Mixing and heat transfer change at scale; pilot control does not carry over

What happens. The process ran well in the pilot plant. In the production unit, mixing is slower, there is relatively less wall to remove the heat, and temperatures react later. The control settings from the pilot oscillate or respond too slowly, and commissioning drags on.

What it costs. Months between mechanical completion and saleable product, batches out of specification during that time, and a team tied up on site. Customers and plans wait.

Why the usual fix fails. Scaling the controller settings by rules of thumb ignores that the dynamics changed, not just the size. Tuning on the large unit by trial and error is slow because every test costs material and time.

How a fixed-price remote diagnosis checks it. I use the pilot data, the design data of the large unit (volume, jacket and coil areas, agitator, cooling capacity) and any commissioning trends to build a simple dynamic model of both scales. The fixed-price remote diagnosis shows where the pilot settings fail on the large unit and whether control designed on the model could shorten commissioning.

Column purity known only from lab samples every few hours

What happens. The distillation column is controlled on tray temperatures, but the real target is product purity, and that comes from the lab a few times per shift. When the result is off, the product of the last hours is already in the tank.

What it costs. Off-spec product to blend, rework or sell cheaper. To avoid it, operators over-purify, which means more reflux and more steam in the reboiler for purity nobody pays for.

Why the usual fix fails. An online analyser is expensive, slow and needs care; many columns do not have one, and where they do, it is often out of service. Tightening the temperature setpoint helps until feed or pressure changes and the temperature no longer means the same purity.

How a fixed-price remote diagnosis checks it. From your historian data (tray temperatures, pressures, flows, reflux, reboiler duty) and the matching lab results, I check how well purity can be computed continuously. The fixed-price remote diagnosis tells you whether such a computed value is good enough to control the column on, and how much reflux and steam margin it could release.

Background: Computed measurements

Feed composition changes faster than the column control can follow

What happens. When the feed to the column changes composition, both products swing for hours. The control system only reacts once tray temperatures have moved, and by then the disturbance is travelling through the whole column. Operators open the margins and wait.

What it costs. Swings in both top and bottom product, off-spec material during every change, and permanently higher reflux and steam to cover the next one.

Why the usual fix fails. Faster controller settings make the column oscillate. A feed analyser is costly and often not available. Asking the upstream unit for steadier feed only moves the problem there.

How a fixed-price remote diagnosis checks it. I look at your upstream data, feed flows, temperatures and lab results to see whether the feed composition can be estimated before it reaches the column. The fixed-price remote diagnosis answers whether a feed estimate combined with predictive control could act before the products swing, and what it would take.

Background: Predictive control of process plants

Reactors run below capacity to keep a margin to runaway

What happens. An exothermic reactor could take more feed, but the temperature control is not trusted near the top. So the feed rate is held well below what cooling could handle, and every increase is discussed shift by shift.

What it costs. Throughput lost every hour the reactor runs below its real limit. In a sold-out market that is lost sales; in any market it is fixed cost spread over less product.

Why the usual fix fails. A faster temperature controller alone does not see the heat release coming. More cooling capacity is an investment. And nobody wants to reduce a safety margin on a hunch.

How a fixed-price remote diagnosis checks it. From your temperature, cooling and feed data, and your safety studies, I estimate how much heat the reaction releases and how much cooling reserve is really left at each load. The fixed-price remote diagnosis shows whether predictive temperature control could run safely closer to the limit, and how that could be demonstrated before anything is changed.

Background: Predictive control of process plants

Conversion and catalyst activity are not measured

What happens. The reactor runs, but how far the reaction goes and how active the catalyst still is are not measured. Operators raise temperature as the catalyst ages, based on lab values and experience. The catalyst change is planned by calendar or forced by falling yield.

What it costs. Yield lost while the catalyst is weaker than assumed, energy spent on temperature the reaction does not need, and catalyst changes at the wrong time, either too early or after weeks of poor output.

Why the usual fix fails. More lab samples cost time and still come late. Online analysers for conversion are expensive and delicate. Fixed temperature schedules ignore that each catalyst charge ages differently.

How a fixed-price remote diagnosis checks it. Temperatures, pressures, flows and lab results together often contain enough to compute conversion and activity continuously; I used this kind of state estimation for the superfluid-helium cooling of the Large Hadron Collider at CERN, the European particle-physics laboratory. The fixed-price remote diagnosis tells you whether your data support it and what it would mean for temperature and change-out planning.

Background: Computing unmeasured states for control

Batches end by the clock, not by their state

What happens. Each batch runs a fixed time because nobody knows exactly when the reaction is complete. The time was set for the slowest batch anyone remembers. Even so, a sample at the end sometimes shows the batch is not finished.

What it costs. Most batches sit in the reactor after they are done, which costs cycle time and energy. The few slow ones come out incomplete and need rework or a second step.

Why the usual fix fails. Shortening the time by hand works until a slow raw-material lot arrives. Extra samples during the batch help, but they keep the operator busy and the results arrive late.

How a fixed-price remote diagnosis checks it. From batch records (temperatures, heating and cooling duty, dosing, pressure, end samples) I check how well the progress of a batch can be computed while it runs. The fixed-price remote diagnosis shows whether the end of each batch could be decided by its computed state, and how much cycle time that would free.

Product changeovers take hours of off-spec material

What happens. Switching the unit from one product to another takes hours. Operators move setpoints in steps they learned over years, wait, take a sample and move again. Everything produced in between is off-spec.

What it costs. Off-spec material to store, blend or sell cheaper, and capacity lost at every change. Planners avoid changeovers, so campaigns grow longer and stocks larger.

Why the usual fix fails. Each operator has a personal recipe for the change, and none is written down in full. Copying the fastest operator’s moves fails when the starting point differs.

How a fixed-price remote diagnosis checks it. I take the historian data of past changeovers and look at which moves shortened the transition and which caused overshoot. The fixed-price remote diagnosis answers whether an optimised transition path, computed from a model of the unit, could shorten changeovers, and what the model would need.

Background: Optimal paths and setpoint plans

Fouling quietly cuts heat recovery in the exchanger network

What happens. Heat exchangers foul slowly. Outlet temperatures drift, the furnace or the steam supply makes up the difference, and nobody notices because control holds the final temperature. Cleaning happens at the next shutdown or when something blocks.

What it costs. Steam and fuel to replace heat that the exchangers should have recovered, week after week. Cleaning at the wrong time costs either production or energy.

Why the usual fix fails. Fixed cleaning intervals ignore that fouling depends on feed and load. Looking at single temperatures does not show it, because flows and inlet conditions change all the time.

How a fixed-price remote diagnosis checks it. From temperatures and flows around the network I compute how the heat transfer of each exchanger develops over months. The fixed-price remote diagnosis shows how much heat recovery fouling costs you today and whether cleaning planned on the computed fouling state would pay.

Evaporator concentration drifts with feed and fouling

What happens. The evaporator should deliver a fixed concentration, but the product drifts when the feed changes or the heating surfaces foul. Density readings or lab samples show the drift late, and operators correct the steam by hand.

What it costs. Product out of specification, more steam to over-concentrate for safety, and cleaning stops that come sooner than planned.

Why the usual fix fails. A standard controller on the outlet density reacts after the fact and struggles with the long delays through several effects. A new density meter shows the drift more clearly but does not prevent it.

How a fixed-price remote diagnosis checks it. From feed flow, temperatures, pressures, steam and outlet measurements I check whether concentration and fouling can be computed through the evaporator. The fixed-price remote diagnosis tells you whether predictive control on a computed concentration could hold the product tighter with less steam, and what it would need.

Crystal size is checked only in the lab

What happens. The cooling profile of the crystalliser is fixed. Whether the crystals came out right is seen only in the lab or at the filter, when the batch is long finished. Fine crystals clog the filter; coarse ones trap impurities.

What it costs. Slow filtration and drying, product rejected for size or purity, and batches reworked. The filter becomes the bottleneck of the plant.

Why the usual fix fails. Changing the cooling profile by trial and error takes many batches, and the result shifts with every new raw material. Inline particle measurement is costly and hard to keep clean.

How a fixed-price remote diagnosis checks it. From temperatures, concentrations, seeding and lab size results of past batches I estimate how supersaturation developed during each batch. The fixed-price remote diagnosis shows whether an optimised cooling profile based on computed supersaturation could give steadier crystal size, and what data it would need.

Compressors are kept far from surge by hand

What happens. To keep the compressor away from surge, the anti-surge valve recycles gas with a wide margin, and operators keep it there because a surge event is feared. The surge line on the drawing is old and nobody trusts it.

What it costs. Gas compressed and recycled for nothing costs power every hour. The margin also limits how far throughput can be turned down or up.

Why the usual fix fails. Narrowing the margin by hand is a risk nobody wants to take. The supplier’s surge line does not account for wear, gas composition or changed operating points.

How a fixed-price remote diagnosis checks it. From pressures, temperatures, flows, speed and valve positions I compute where the compressor really operates relative to surge. The fixed-price remote diagnosis shows how much margin is really needed and whether control on a computed surge distance could safely run closer to the line.

Additives are dosed high to be on the safe side

What happens. Inhibitors, neutralising agents and other additives are dosed at a fixed rate that covers the worst feed. Nobody measures how much is really needed at the moment.

What it costs. Chemicals paid for and then wasted, and sometimes side effects in the product or the wastewater. On a large unit the overdosing adds up to a steady cost.

Why the usual fix fails. Reducing the dose by hand is tried until one bad incident, then it goes back up. A dosing table by product grade ignores how much the feed varies within a grade.

How a fixed-price remote diagnosis checks it. I look at dosing rates, feed data, process conditions and the lab results that show the effect of the additive. The fixed-price remote diagnosis answers whether the real demand can be estimated from what you measure, and how much additive a dose tied to that estimate could save.

When the analyser fails, the loop falls back to manual

What happens. An online analyser feeds a key control loop. When it drifts, calibrates or fails, the loop goes to manual and operators hold the unit by feel until the technician has repaired it. That can take days.

What it costs. Off-spec product during the outage, operators tied up with one loop, and margins opened to be safe. Analyser maintenance itself is expensive and specialists are scarce.

Why the usual fix fails. A second analyser doubles the cost and often fails for the same reasons. Holding the last good value works for minutes, not for days.

How a fixed-price remote diagnosis checks it. From analyser readings, the process measurements around them and maintenance logs I check how well the analysed value can be computed from other signals, and how a drifting analyser could be spotted early. The fixed-price remote diagnosis shows whether a computed value could keep the loop in automatic during outages.

Operators drown in alarms during upsets

What happens. When something goes wrong, hundreds of alarms come in within minutes. The first, important one is buried among the follow-on alarms. Operators silence them and work from experience.

What it costs. Wrong or late actions in exactly the situations where they matter most, longer upsets, and a real safety risk. New operators are overwhelmed.

Why the usual fix fails. Alarm rationalisation removes the worst nuisance alarms, but the flood during an upset comes from real follow-on effects. Suppressing them by fixed rules hides information when the situation differs from the rule.

How a fixed-price remote diagnosis checks it. From alarm and event logs of past upsets and the process data around them, I trace which first deviation led to which cascade. The fixed-price remote diagnosis shows whether a model of the unit could point operators to the likely root cause during an upset, and for which kinds of upset.

Every start-up needs the most experienced shift

What happens. Starting the unit up after a shutdown is done by hand, and only a few people know how. Start-ups are scheduled around their shifts. Each one is a little different, and it takes hours before the product is in specification.

What it costs. Off-spec material and lost hours at every start-up, risk while the plant passes through unusual states, and a dependency on people who will retire.

Why the usual fix fails. Start-up procedures on paper describe the steps but not the timing and the judgement. Automating the sequence with fixed steps breaks when conditions differ from last time.

How a fixed-price remote diagnosis checks it. I compare the historian data of several past start-ups: what was moved when, what went smoothly, where time was lost. The fixed-price remote diagnosis answers whether a model-based start-up sequence, or guidance for the operator, could make start-ups shorter and less dependent on who is on shift.

Process changes cannot be tried without risking the plant

What happens. Engineers have ideas: a different control strategy, a new operating point, a changed sequence. But the only place to try them is the running plant, and nobody wants to risk production on an idea. So most ideas are never tested.

What it costs. Improvements left unused year after year, and changes that are made anyway sometimes cause upsets that a test would have shown.

Why the usual fix fails. Steady-state design models show where the plant ends up, not how it gets there. A full simulator of the whole site is a large project that rarely pays for a single question.

How a fixed-price remote diagnosis checks it. I look at the process data and documentation of the unit in question and judge whether a dynamic model of just that part can be built from what you have. The fixed-price remote diagnosis shows which of your open ideas such a model could test first, and what it would take to build.

Particle size of emulsion polymers is known only after the batch

What happens. In emulsion polymerisation the particle size decides the properties of the product, but it is measured in the lab after the batch. During the batch, operators follow the recipe and hope.

What it costs. Batches out of specification are reworked, blended or downgraded. Raw-material lots that behave differently are discovered only when the result is wrong.

Why the usual fix fails. Tighter recipe control does not help when the raw materials vary. Inline particle-size measurement is costly and fouls quickly in a latex.

How a fixed-price remote diagnosis checks it. From dosing, temperatures, heat release and lab results of past batches, I check how well particle growth can be computed during the batch. The fixed-price remote diagnosis shows whether such a computed value could warn early or steer dosing, and how close it gets to the lab.

Advanced control is wanted at every site, but too few people can build it

What happens. Many chemical groups want advanced process control and online optimisation on every plant, not only on the flagship one. The plan exists; the people to carry it out do not. Each site waits its turn while the few specialists move from project to project.

What it costs. Plants run for years on standard controllers and wide margins while the benefit waits. Projects that start lose momentum when the specialist is pulled away.

Why the usual fix fails. Hiring control engineers takes a long time and they are scarce. Large vendor packages need internal people to maintain them, and they get switched off when nobody looks after them.

How a fixed-price remote diagnosis checks it. For one unit I look at the data, the control structure and the operating targets and judge where predictive control or optimisation would pay most and what it needs. The fixed-price remote diagnosis gives you a clear picture for one site, which helps decide where to start and what an outside engineer can carry.

Background: Economic predictive control

The grid-fee discount for steady consumption ends; flexibility will pay instead

What happens. In Germany, large consumers with a very steady load have long paid strongly reduced grid fees. The conditions are at least ten gigawatt hours a year and over 7,000 full-load hours, and the discount can reach up to 90 % (innobu). The regulator is replacing this privilege; the new system no longer rewards a flat load.

What it costs. A plant built to run flat out may lose a discount that is a large share of its grid costs, while what the new rules reward is load it cannot yet shift.

Why the usual fix fails. Simply running the plant down when power is expensive costs product and stability. Buying batteries is expensive when the plant already has buffers it does not use.

How a fixed-price remote diagnosis checks it. From your load profile, production plan, tank levels and storage I check which loads can shift and by how much, without losing output or quality. The fixed-price remote diagnosis shows whether a schedule computed from a model of the plant could use that flexibility, and what it would take.

Background: Economic predictive control

Industrial power price relief comes with decarbonisation measures to prove

What happens. Germany’s subsidised industrial power price lowers the bill for energy-intensive firms, but part of the aid has to go into measures that reduce emissions, within a fixed period. Firms that cannot show such measures may have to pay the aid back.

What it costs. Engineering teams need investable measures with savings they can prove, quickly, on top of normal work. Measures chosen in a hurry may not deliver.

Why the usual fix fails. Lists of standard measures (new motors, insulation, heat pumps) are quickly drawn up but say little about the real savings in your process. Energy audits name candidates, not figures you can defend.

How a fixed-price remote diagnosis checks it. From your energy data, process logs and the measures already under discussion, I check which ones a model of the plant shows to save the most, and how the savings could be measured afterwards. The fixed-price remote diagnosis gives you a ranking you can argue for, with the assumptions behind it.

Energy management systems collect data that nobody uses to run the plant

What happens. German companies with an average final energy consumption of more than 7.5 gigawatt hours a year must have an energy or environmental management system (Energy Efficiency Act, section eight). The meters are installed and the reports are written.

What it costs. The data cost money to collect and certify, but they rarely change how the plant runs. Savings stay on paper, and the next audit finds the same things.

Why the usual fix fails. Monthly energy reports are too coarse to show what drives consumption shift by shift. Dashboards show numbers without saying which setpoint to change.

How a fixed-price remote diagnosis checks it. I take the meter data from your management system together with process data and look at what really drives energy use in each unit. The fixed-price remote diagnosis shows whether the data you already collect can feed computed values and better control, and where that would save most.

Plants run far below design load with controllers tuned for full load

What happens. Many chemical plants remain under-utilised: production in the first half of 2026 was about 3 % below the year before, and investment in fixed assets is about 15 % below 2023 (CHEManager). The controllers were set up for full load.

What it costs. At part load, loops become sluggish or oscillate, columns and furnaces run inefficiently, and energy per tonne rises. Turned-down units often need more operator attention, not less.

Why the usual fix fails. Retuning for one load fails at the next. Running at full load and then stopping is not always possible, and new equipment for part load is not in the budget.

How a fixed-price remote diagnosis checks it. From historian data at different loads I check which loops and units lose most at turndown. The fixed-price remote diagnosis answers whether retuning for the real load range and an optimised part-load operating point could recover efficiency without new equipment.

Controller migrations open control code nobody has touched for years

What happens. Many plants must move from older controller generations to new ones before the old hardware runs out of support. Not everything converts automatically, so the control code is opened and rewritten, often by people who did not write it.

What it costs. Engineering time, a shutdown for the switchover, and the risk that loops which worked by accident stop working. Old workarounds are copied into the new system without anyone knowing why they are there.

Why the usual fix fails. A one-to-one migration preserves the old hand-tuned loops with all their faults. Testing the new code on the live plant is risky.

How a fixed-price remote diagnosis checks it. From the existing control code, loop data and operating history I check which loops perform poorly today and why. The fixed-price remote diagnosis shows where the migration is a chance to replace hand-tuned loops with model-based control and documented settings, and what that would add to the project.

Stricter industrial emissions rules need performance data from operation

What happens. The revised European industrial emissions rules are being implemented in German law. Permitted plants face stricter requirements based on the best available techniques, and existing plants will need an environmental management system that tracks their performance.

What it costs. Tighter emission levels must be held in daily operation, and the management system needs reliable performance data, not estimates assembled once a year.

Why the usual fix fails. More analysers cover the measured points, not the whole picture. Holding new limits with a wider margin costs energy and reagent every day.

How a fixed-price remote diagnosis checks it. From your emission measurements, process data and permit conditions I check where the plant is close to the new levels and which values could be computed to fill the gaps. The fixed-price remote diagnosis shows whether computed values and better control could hold tighter limits and feed the management system from operation.

Swiss target agreements: the promised savings measures must be delivered

What happens. In Switzerland, large consumers can be exempted from the carbon levy by signing long-term target agreements recognised by the federal government. The agreement lists economic measures the company commits to carry out over the years.

What it costs. Each measure on the list has to be delivered and has to pay back within the expected time. Missing the target puts the levy exemption at risk, and the engineering effort comes on top of daily work.

Why the usual fix fails. The obvious measures, lighting, insulation, compressed-air leaks, are often done already. The remaining potential lies in the process itself, where standard energy advice reaches its limit.

How a fixed-price remote diagnosis checks it. From process data, energy meters and the agreed measures I check where process control and operating optimisation could save energy with a short payback. The fixed-price remote diagnosis tells you which process measures are realistic and what they would take.

Energy audits ask for savings that can be measured

What happens. European energy efficiency rules oblige large energy users to carry out regular energy audits or to run an energy management system.

What it costs. Firms have to show savings they can measure, not intentions. Audits that list measures without follow-up lead to the same findings the next time, and management asks why nothing changed.

Why the usual fix fails. Audit reports name generic measures. Implementing them in a running process without a model of the plant is guesswork, and the saving is hard to separate from changes in load and product mix.

How a fixed-price remote diagnosis checks it. From your audit findings, energy meters and process data I check which process measures would save most and how the saving would be measured against load and product mix. The fixed-price remote diagnosis gives you measures with a computed effect.

Climate protection contracts: new plants must meet their reduction milestones

What happens. Germany’s climate protection contracts are open to industrial steam and carbon capture projects, and they come with emission reduction targets the project has to reach over the years (Chamber of Industry and Commerce Bonn).

What it costs. New plant must reach its reduction milestones on time. A process that runs below its design performance puts the payments at risk.

Why the usual fix fails. Commissioning teams get new equipment running, but optimising it for emissions and cost is a different job, often left for later, when the first milestone is already close.

How a fixed-price remote diagnosis checks it. From the design data, commissioning trends and the agreed milestones I check where the new process is likely to miss its targets and why. The fixed-price remote diagnosis shows whether control and optimisation could close the gap, and what that would take.

Funded low-carbon projects must build fast and know their cost of heat

What happens. The European Innovation Fund supports highly innovative low-carbon projects, and separate auctions for industrial heat are planned (PNO Innovation). Winners must build quickly; heat bidders must know what a unit of heat really costs them.

What it costs. Delays in engineering and commissioning eat into funded timelines. A bid with a wrong heat cost either loses or loses money.

Why the usual fix fails. Spreadsheet estimates of heat cost assume steady full load. They ignore how the plant really runs: part load, start-ups, swings in power price, storage that fills and empties.

How a fixed-price remote diagnosis checks it. From your design data, load profiles and energy prices I build a first dynamic model of the heat supply and check how the cost per unit of heat varies with operation. The fixed-price remote diagnosis tells you whether your heat-cost figure holds and where control could lower it.

Operating know-how retires with the people who carry it

What happens. The operators and engineers who know how the plant behaves in odd situations are retiring. Much of what they know is not in any document; it is in how they react to a trend.

What it costs. Upsets last longer, start-ups take more time and quality varies more once they are gone. Younger staff learn by trial and error on the real plant.

Why the usual fix fails. Interviews and handover documents capture rules, not judgement. Keeping retirees on contract only delays the problem by a few years.

How a fixed-price remote diagnosis checks it. From historian data of the situations your experts handle best, I check which of their decisions can be explained by the process data. The fixed-price remote diagnosis shows whether that know-how can be turned into models, computed values and control logic that stay in the plant.

Lab results arrive hours after the process has moved on

What happens. Key quality values come from the lab hours after sampling, or from analysers that need constant maintenance. By the time a result arrives, the process has moved and the correction is too late.

What it costs. Off-spec material produced while waiting, wide margins to compensate, and analyser technicians kept busy.

Why the usual fix fails. More frequent sampling costs lab time and still arrives late. More analysers mean more maintenance. Lab and process data are often kept in separate systems, so nobody links them.

How a fixed-price remote diagnosis checks it. I bring lab results and process data together and check which quality values can be computed continuously from pressures, temperatures and flows. The fixed-price remote diagnosis shows how computed values and analysers could back each other up, and which units would gain most.

Pumps, fans and compressors run against throttles

What happens. Electric motors take about 27 % of Switzerland’s total power use, and too many still run with poor efficiency (SRF).

What it costs. Pumps and fans that run at full speed against a throttle valve or damper waste power every hour. Peak loads drive up grid charges.

Why the usual fix fails. Replacing motors with better ones helps, but the larger loss is often in how the drive is operated. Variable-speed drives installed without changed control still run near full speed.

How a fixed-price remote diagnosis checks it. From power meters, valve positions, speeds and flows I check where drives waste power and where the peaks come from. The fixed-price remote diagnosis shows which changes in drive and process control would cut power use and peaks, and in what order.

Emissions trading: steam above the benchmark now costs allowances

What happens. The European Commission adopted new benchmarks for free allocation in emissions trading on 29 June 2026. Free allocation is expected to fall by more than 16 % on average compared with 2021, and the heat and fuel fallback benchmarks by 34 % each (Refindustry).

What it costs. Every tonne of steam or fuel above the benchmark now has to be covered with bought allowances. Steam systems that were good enough become a cost line.

Why the usual fix fails. Boiler upgrades take years. Cutting steam by hand risks the process units that depend on it.

How a fixed-price remote diagnosis checks it. From steam balances, boiler data and the consumption of each unit I check where steam and heat are lost and which units use more than they need. The fixed-price remote diagnosis shows where steam optimisation, heat recovery and better control would cut allowance costs.