The hard part is already built
A training simulator needs a plant model that responds correctly to operator action. That is exactly what the reference model does, and it is calibrated against a real train rather than against a generic flowsheet.
Operator Training Simulator
OTS
Train the crew on the faults they have never seen.
A reference model that knows what healthy looks like, and a fault library that knows what twenty-two failures look like, is most of a training simulator already. OTS turns that into a rig: inject a fault, let the trainee respond, and score what they did in money rather than in pass or fail.
In devCurrent status
22Faults available to inject
6Cost buckets for scoring
1:1Same screens as operations
Why this is next
Most of this is already built. It has not been assembled yet.
A fault library, not a scripted scenario
Twenty-two signatures already carry their evidence, their progression and their causal parents. Scenarios can be built from recorded episodes that genuinely happened on the unit, rather than written by hand.
Scored in money
The consequence model already prices off-spec gas, amine loss, excess steam, reliability loss and lost throughput. A trainee's decisions can be scored on what they actually cost, which is a far better teacher than a mark out of ten.
The same screens as the real job
Trainees work on the marked P&ID that operations uses. There is no transfer gap between the simulator and the control room.
Planned scope
What we intend to build.
This is a plan, not a product. Nothing in this list is shipping today.
- Scenario library assembled from recorded fault episodes on the unit
- Fault injection with adjustable severity and ramp rate
- Free-running plant response to trainee actions, not a fixed script
- Scoring against the consequence cost model
- Instructor mode — pause, rewind and branch a session
- Per-operator session reports and competency tracking
What it builds on
It reuses what is already calibrated.
Every figure on this page is countable in the delivered system. If you want to see where a number comes from, ask us and we will show you the record it is read from.
- Built on
- The Run reference model and the Detect fault library
- Scenario source
- Recorded fault episodes, plus synthetic severity ramps
- Scoring
- The same six consequence buckets used in live diagnosis
- Interface
- The marked P&ID used by operations
- First target
- Amine gas sweetening and post-combustion capture
Where it stands today
The limits, published on the same page as the promise.
- This is in development. We do not publish a delivery date we have not committed to a customer.
- The first release will cover the same unit family as the rest of the platform. It is not a general-purpose flowsheet simulator and it is not intended to replace one.
- A simulator is only as good as the model behind it. Ours is calibrated to one train, so the honest scope today is training on that unit family — not on any plant.
Where it applies
Built for amine systems today. Not limited to them by design.
The platform is focused on acid gas removal — gas sweetening and post-combustion carbon capture — because that is where the model is calibrated and where the fault library is real. The method underneath is not amine-specific: a physics reference model, residuals against expected-healthy, fuzzy diagnosis over graded indicators, and costed consequences. Extending it to another unit means new datasheets and a new calibration, not a new platform.
The rest of the platform
One model. Seven ways to use it.
The fastest way to judge this is to open it.
The platform runs on a real amine train with thirteen days of one-minute history behind it. Nothing on these pages is a mock-up.
Open the platform