13. Sample Plants and Scenario Packs¶
The course so far used one plant. Real training needs more than one: a trainee who has only ever seen a separator has not learned much about coolers, drums or cascades of vessels. This page describes the plants that ship with the OTS, each saved with its screens, interlocks and a pack of scored scenarios, ready to load and run.
What you will learn
- What each sample plant contains: tags, loops, alarm limits, interlocks
- The scenarios in each pack, what they do to the plant and what the trainee has to achieve
- Where the numbers come from, so you can adjust them
Prerequisites
- Page 12: you know how a session runs from briefing to debriefing
The three plants¶
| File | Plant | Loops | Scenarios | Best for |
|---|---|---|---|---|
ots_separator.dwxmz |
Gas-liquid separator (pages 1 to 12) | Pressure, level | The one built on page 4, plus whatever you add | First contact with the OTS |
ots_two_stage_separation.dwxmz |
HP and LP separators in series | 2 pressure, 2 level | 7 | Level and pressure interaction, gas blow-by, trips |
ots_aftercooler.dwxmz |
Compressor aftercooler and knock-out drum | Flow, temperature, level, pressure | 7 | Heat exchange, instrument failures, utility loss |
Every file has the initial state a stored and selected in schedule1, a 5 s step (2 s for the aftercooler), the historian on, an Overview screen with detail screens, and the interlocks listed below. Open the file, open the Instructor Station, pick a scenario, Load, Start.
Each pack follows the same progression: scenario 1 has no fault and teaches the plant; 2 to 5 have one fault each, in order of difficulty; 6 is a plant-wide failure; 7 is an assessment at real time.
Two-stage separation¶
A rich gas from the wells enters at 45 bar through FV-100 into the HP separator V-100 (6 m³, 4 m tall), held at 30 bar by PIC-100 on the gas valve PV-100 to a 25 bar header. The HP condensate drops through LV-100 (level loop LIC-100, set point 1.2 m) to the LP separator V-200 (3 m³, 3 m tall) at 8 bar (PIC-200 on PV-200, to a 3 bar header); its liquid leaves through LV-200 (LIC-200, 1.0 m) to a 1.5 bar header. Feed 3 kg/s: 0.32 kg/s of HP gas, 0.19 kg/s of LP gas and 2.5 kg/s of stabilised condensate. Peng-Robinson, seven hydrocarbons from methane to n-heptane.
The level controller of V-100 reads a transmitter, LT-100, and the gauge LIT-100 reads the true level, so the sensor scenarios work the way page 3 describes.
| Tag | Reads | Units | LL | L | H | HH |
|---|---|---|---|---|---|---|
| PIT-100 | V-100 pressure | bar | 20 | 26 | 34 | 37 |
| LIT-100, LT-100 | V-100 level | m | 0.4 | 0.7 | 1.8 | 2.2 |
| PIT-200 | V-200 pressure | bar | 4 | 6 | 10.5 | 12 |
| LIT-200 | V-200 level | m | 0.3 | 0.5 | 1.6 | 2.0 |
| FI-100, FI-101, FI-200, FI-201 | Feed, HP gas, LP gas, condensate | kg/h | ||||
| TI-100 | Feed temperature | °C |
Interlocks:
| Name | Trips when | Actions |
|---|---|---|
| V-100 low level trip | LIT-100 LL for 3 s | Close LV-100; LIC-100 to manual at 0 % (gas blow-by protection) |
| V-200 high pressure trip | PIT-200 HH for 3 s | Close LV-100; PIC-200 to manual at 100 % |
| V-100 high pressure trip | PIT-100 HH for 5 s | Close FV-100 |
Two schedules besides the plain one carry a disturbance as an event: schedule2 (feed surge) steps the feed header from 45 to 50 bar at 01:30, schedule3 (large feed surge) to 60 bar. Scenarios pick the schedule they need.
| # | Scenario | Speed | Fault | What happens without action | Objectives |
|---|---|---|---|---|---|
| 1 | Reading the plant | 5x | none | Nothing; the trainee changes the PIC-100 set point and watches | Keep PIT-100 in 26 to 33 bar and LIT-100 in 0.7 to 1.8 m; change the set point within 10 min |
| 2 | LV-100 sticks during a feed surge | 10x | LV-100 stuck at 01:00; surge at 01:30 | Level H at 17:05, HH at 27:00 | No HH; LIC-100 to manual within 25 min; level in range 80 % of the time |
| 3 | Gas blow-by | 5x | LT-100 fails high at 01:30: LIC-100 opens LV-100 wide | True level L at 04:50, LL at 06:35, low level trip at 06:40; after the trip the level climbs to HH at 17:15 | No V-200 HH; no low level trip; LIC-100 to manual within 7 min; level in 0.7 to 1.8 m 70 % of the time |
| 4 | PV-100 sticks during a feed surge | 2x | PV-100 stuck at 01:00; surge to 60 bar at 01:30 | Pressure H at 09:00, settles near 35 bar | Pressure in 26 to 33 bar 70 % of the time; no HH; no high pressure trip; throttle FV-100 within 5 min |
| 5 | Frozen level transmitter | 10x | LT-100 frozen at 01:00; surge at 01:30 | LIC-100 holds its output; true level H at 17:05, HH at 27:00 | No HH, no LL; LIC-100 to manual within 30 min; level in range 80 % |
| 6 | Instrument air failure | 5x | All valves to fail position at 03:00 | Feed stops, everything holds | Acknowledge within 5 min; no HH pressures |
| 7 | PV-100 fails closed (real time) | 1x | PV-100 closes at 02:00 over 3 min | Pressure H at 05:25, HH at 07:00, trip at 07:05 | No HH; no trip; throttle FV-100 within 5 min |
The times in the table are what the plant does with nobody at the panel; they were measured by running each scenario headlessly. They tell you how long the trainee has: in scenario 3, a hundred and five seconds of simulated time between the L and the LL alarm, twenty-one seconds of wall-clock time at 5x. Slow the scenario down if that is too little for a beginner.
Aftercooler and knock-out drum¶
Compressed gas at 10 bar and 150 °C enters through FV-101 (flow loop FIC-101, 5 400 kg/h) into the aftercooler E-101, leaves through the hand valve XV-101 into the knock-out drum D-101 (8 m³, 4 m tall) at 7 bar (PIC-101 on PV-101 to a 6 bar header); the condensate, about 0.4 kg/s of the heavier ends, goes through LV-101 (LIC-101, 1.0 m) to a 3 bar header. The cooling duty of E-101 is what TIC-101 sets from the gas outlet temperature (set point 45 °C); the cooling water is behind it. The temperature controller reads the transmitter TT-101; the gauge TI-101 reads the true temperature. This plant integrates with a 2 s step.
| Tag | Reads | Units | LL | L | H | HH |
|---|---|---|---|---|---|---|
| TI-101, TT-101 | Gas outlet temperature | °C | 30 | 35 | 60 | 75 |
| FI-101 | Gas flow | kg/h | 1500 | 3000 | 9000 | 10000 |
| LIT-101 | D-101 level | m | 0.3 | 0.5 | 1.25 | 1.4 |
| PIT-101 | D-101 pressure | bar | 5 | 6 | 8 | 9 |
| TI-100, TI-103, QI-101, FI-103, FI-104 | Gas inlet T, drum inlet T, cooling duty, condensate, gas to header |
Interlocks: Hot gas to drum (TI-101 HH for 10 s: close FV-101, FIC-101 to manual at 0), D-101 high pressure trip (PIT-101 HH for 5 s: close FV-101, PIC-101 to manual at 100 %), D-101 overfill (LIT-101 HH for 5 s: close FV-101). The schedule schedule2 (more gas) raises the FIC-101 set point from 5 400 to 8 500 kg/h at 01:30.
| # | Scenario | Speed | Fault | Objectives |
|---|---|---|---|---|
| 1 | Reading the plant | 5x | none | Keep TI-101 in 35 to 60 °C and the level in 0.5 to 1.25 m; change the TIC-101 set point |
| 2 | Cooling water shortfall | 5x | E-101 loses 40 % of its capacity over 10 min from 01:00; H alarm at 06:38 | No HH temperature; temperature in range 80 %; reduce the FIC-101 set point |
| 3 | Frozen temperature transmitter | 5x | TT-101 frozen at 01:00, more gas at 01:30: TIC-101 keeps the cooling where it was; H alarm at 02:02, no HH | No HH; TIC-101 to manual within 15 min; temperature in range 70 % |
| 4 | LV-101 sticks | 10x | LV-101 stuck at 01:00, more gas at 01:30; the level climbs about 0.85 cm/min, H at 29:20, HH at about 47 min | No overfill trip; no HH level; LIC-101 to manual; reduce the gas |
| 5 | Drifting temperature transmitter | 5x | TT-101 drifts high from 02:00: TIC-101 over-cools | No LL temperature; TIC-101 to manual; temperature in range 80 % |
| 6 | Instrument air failure | 5x | All valves to fail position at 03:00 | Acknowledge; no HH temperature or pressure |
| 7 | PV-101 fails closed (real time) | 1x | PV-101 closes at 02:00 over 4 min; H at 04:02, HH at 04:54, trip at 05:00 | No HH pressure; no trip; cut the gas within 5 min |
How the plants were built¶
Both files were generated with the Fluent API rather than drawn by hand, so that every number is reproducible: the steady state is solved first, each control valve is then given a Kv that puts it at 50 % open at the design flow, the plant is run under the OTS session for an hour of simulated time to settle, and the settled state is stored as a. The scenarios, interlocks and screens are written into the file with the same classes the Instructor Station uses. Page 16 shows how to drive those classes from Python.
Two design choices are worth copying when you build your own plant:
- Vessel size sets the pace. Pressure in a gas space moves in seconds, level in a liquid pool moves in minutes. The HP separator is tall and narrow (4 m for 6 m³) so that a 0.4 kg/s liquid imbalance moves the level a few centimetres per minute, fast enough to see and slow enough to act on. Scenario speeds are chosen per scenario: 2x for the pressure exercise, 10x for the level ones.
- A stuck valve at steady state is invisible. Nothing changes until something else changes. The stuck-valve scenarios pair the fault with a disturbance carried by the schedule's event list (a feed surge, hotter gas), which is what the extra schedules are for.
The next page gives you the paperwork for running these plants with a class.