A steam plant knows how much fuel it burns. What it cannot see is how much useful heat that steam actually delivers. PhaseVector measures it from the sensors already on the plant, and states how much to trust the answer.
Physics-informed AI with a machine-learning layer trained on the residual.
Measurement assurance
A stated confidence band and explainable reasoning on every output.
Industrial fit
Reads the historian or SCADA a plant already runs, cloud or on-premise.
The blind spot
Gauges confirm the steam is there. They never say how wet it is.
On saturated steam, pressure and temperature move together, so they confirm steam is present without revealing its moisture. Real steam carries three to seven percent. Every flow computer assumes it carries none, and because the error is systematic it never shows up as an anomaly in anyone's reporting.
Delivered energy at eight bar, steam five percent wet
What the flow computer logsassumes perfectly dry
What is really delivered3.7% lower
3.7%
overstated on a total-enthalpy basis, and 5.3% counting only the latent heat the process actually uses.
It is thermodynamics, not calibration. No amount of instrument accuracy recovers a quantity the measurement does not contain.
It biases the decisions built on it. Efficiency figures, carbon reporting, project savings and electrification sizing all inherit the same error.
Specialist instruments exist, and prove the market values the answer. All are capital equipment fitted to a single pipe, needing purchase approval and usually a shutdown.
PhaseVector closes it in software. Continuous and site-wide, wherever the tags already exist, with no equipment vendor's interest in the answer.
The mis-statement38,000 tCO2e
a year across Scotch whisky, reported in good faith against an assumption nobody has checked.
The prize£75k to £112k
recoverable a year at a mid-size distillery burning roughly 20 GWh, per US Department of Energy findings of 10 to 15 percent.
The carbon366 to 549 t
of CO2e a year at that same site, verified continuously rather than once, so a saving that drifts back is caught.
Why this matters
Conventional steam monitoring shows flow, but not the full truth behind it.
When steam quality drifts, operators face over-reading, weaker heat transfer, unstable conditions, and shaky audit evidence. PhaseVector adds an intelligence layer that turns raw plant signals into earlier, more actionable insight.
01
Monitor
Steam quality and delivered energy at every monitored point.
02
Diagnose
Hidden losses and non-obvious performance drift, as they emerge.
03
Validate
Energy, carbon, and audit evidence with the reasoning attached.
04
Scale
From one steam main to a whole estate, without new hardware.
Steam focus
Wet steam. Hidden losses. Earlier action.
Detect instability before it distorts energy accounting, process efficiency, and decarbonisation planning.
Scotch whisky alone burns around 6,700 GWh of heat a year, roughly five percent of Scotland's energy, and about 83 percent of that goes into raising steam. For breweries and distilleries this turns steam quality into audit-grade evidence: stronger SECR reporting, defensible SWA net-zero baselines, and a correctable view of Scope 1 emissions as sites decarbonise. When the delivered-energy figure is trustworthy, so is every claim built on it.
SECR reportingSWA net zero 2040Scope 1 decarbonisationEnergy auditsGrant M&VElectrification sizing
The value
What you lose without it, and what you gain with it.
Steam quality quietly shapes cost, efficiency, and the numbers you report. Making it measurable turns a hidden liability into a defensible advantage.
Without measurement intelligence
Delivered energy overstated by around 3.7 percent, and 5.3 percent on latent heat alone
Recoverable losses a steam survey would find stay uncaptured between surveys
Carbon and efficiency figures reported in good faith against an unchecked assumption
Electrification and heat-pump projects sized against a baseline that is wrong by around four percent
No way to tell a failing instrument from a genuinely changed process
With PhaseVector
A continuous steam quality and delivered-energy figure with a stated confidence band
Recoverable losses found continuously rather than once, so savings that drift back are caught
Audit-ready carbon and energy evidence, traceable from sensor to report
A measured baseline that makes electrification business cases financeable and verifiable
Automatic drift detection and recalibration, with a signal-quality gate before go-live
The platform
Designed like a deep-tech industrial product, not a generic dashboard.
A subscription software platform, delivered from the cloud or on-premise. It reads the pressure, temperature and flow signals a plant already logs, applies a physics model of the steam network with a machine-learning layer trained on the residual, and returns a continuous estimate of steam quality and delivered energy at every monitored point.
Predictive steam diagnostics
Flag anomalous steam behaviour early from historical and live data, before it grows into an energy or maintenance problem.
Confidence you can check
Every output carries a stated uncertainty band and shows which signals drove it, so an engineer can interrogate the number before acting on it.
Built for degraded data
Automatic drift detection and recalibration, with a signal-quality go or no-go gate at setup, so the platform never becomes confidently wrong.
How an engagement runs
1
Discovery
Map steam assets, data availability, existing instrumentation, and site priorities.
2
Baseline
Connect to the historian, calibrate to that plant, and establish operating signatures and confidence limits.
3
Validation
Benchmark against a traceable reference so the accuracy claim is independently confirmed.
4
Operational use
Turn validated insight into alerts, reporting evidence, and reliability improvement.
Where it applies
One measurement-intelligence core, portable without redesign.
The same physics problem exists wherever saturated or wet steam carries energy, which makes the platform portable across process industries running comparable steam networks under the same reporting obligations.
Mash tun, wash and spirit stills, and the main header, where uniform latent heat transfer decides both product and fuel bill.
Beachhead
Breweries
Wort boiling, pasteurisation, cleaning and utilities, with steam quality tracked as a process-confidence signal.
Beachhead
Dairy & food
Comparable steam networks under the same statutory reporting obligations, with sterilisation temperatures that depend on dry steam.
Pharmaceuticals
Clean and process steam where quality is a compliance question as much as an energy one.
Paper & chemicals
Large continuous steam loads where a systematic energy error compounds across every drying and reaction stage.
Power generation
Cycle and turbine performance, where wet steam quietly shapes efficiency and reliability programmes.
Hydrogen & gas systems
Energy systems where robust diagnostics and trusted flow data decide what gets built and what gets billed.
Geothermal
Most wells deliver a two-phase mixture at the wellhead, and its energy content determines field output.
Adjacent
Making industrial measurement more intelligent, trusted, and actionable, helping operators make better decisions with greater confidence, efficiency, and reliability.
PhaseVector was founded by a specialist in multiphase flow measurement and physics-informed machine learning, with a career across oil and gas, the energy transition, and AI for complex industrial measurement systems.
The platform commercialises a precise method class: machine-learning models that state their own confidence level, developed for gas flow measurement through a joint industrial doctorate and extensive high-pressure experimental testing, and applied here to steam. Scientific advisors cover multiphase flow and digital twins, and chemical thermodynamics and the energy transition.
Multiphase flow meteringCFDPhysics-informed AIUncertainty quantification20+ publications300+ citationsUK Global Talent
"PhaseVector is the independent referee for industrial steam energy."
Software, site-wide, with no equipment to sell and no interest in the answer.
Contact
Book a discovery call.
Start with a short technical conversation about your steam system, the data you already log, and what a measured baseline would be worth. Enquiries come straight to us by email.
This notice explains how PhaseVector Technologies handles personal data collected through this website, in line with UK GDPR and the Data Protection Act 2018.
Who we are
PhaseVector Technologies, Glasgow, United Kingdom, is the data controller for information submitted through this site. Contact: hosseinisa@phasevector.co.uk.
What we collect
Only what you choose to enter in the enquiry form: your name, role, organisation, email address, site type, whether you operate a historian or SCADA system, and your message. We do not collect plant data through this website.
Why we use it, and our lawful basis
We use it solely to respond to your enquiry and to discuss a possible engagement. Our lawful basis is legitimate interests, namely replying to a business enquiry you have initiated. We do not use it for marketing lists, and we do not sell or share it for advertising.
How it reaches us
The form is delivered to our inbox by FormSubmit, a third-party form-delivery service acting as a processor, and is then stored in our business email system. Both may process the message in transit. No other third party receives it.
How long we keep it
Enquiries are kept for up to 24 months from our last contact with you, then deleted, unless we have entered into an agreement that requires us to keep records for longer.
Confidentiality of technical information
Please do not send confidential plant data through this form. Where a discussion progresses, we sign a non-disclosure agreement before any operational data is shared, and plant data remains yours.
Cookies and analytics
This site sets no cookies and runs no analytics or tracking. Fonts are loaded from Google Fonts, which means your browser makes a request to Google and your IP address is visible to them as part of serving those files.
Your rights
You have the right to request access to your data, to have it corrected or erased, to restrict or object to its processing, and to data portability. Email us and we will respond within one month. If you are unhappy with our response, you may complain to the Information Commissioner's Office at ico.org.uk.
