Probabilistic Digital Twin Dashboard

Pipeline Structural Operation and Maintenance

DNV, 2018. Mobile demonstration of a probabilistic digital twin for pipeline over-pressure risk.

Burst probability 2.60e-3 Target 1.00e-3
Pipeline pressure 165 bar MAOP 200 bar
Volume flow 14.3 MSm3/d Steady export
HIPPS reliability 0.936 Next proof test 34 d
Risk status Warning Above target

Live Model Stage

Layer 2: Digital Twin

The physical system is in steady export. Sensor evidence is mirrored by the process model.

  1. Physical system
  2. Digital twin
  3. GP surrogate
  4. Probabilistic DT

1. Physical Pipeline System

Offshore export pipeline with PCS and HIPPS

150 km
Offshore Platform Compressor SSIV HIPPS open 100% Blockage Gas Treatment Plant Subsea Gas Export Pipeline HIPPS 230 bar PCS 210 bar MAOP 200 bar PT / FT / TT 165 bar PT / FT / TT 14.3 MSm3/d

PCS and HIPPS are protective barriers. They reduce or isolate pressure demand; they do not strengthen the pipe material.

2. Gaussian Process Surrogate

3D pressure surface

Pwr 83 MW / Pout 95 bar

Drag the surface to rotate. X = compressor power, Y = outlet backpressure, Z = pipeline pressure.

3. Probabilistic Digital Twin

Bayesian network

Conditional risk
Flow process model Safety system reliability model Pipe structural reliability model Compressor 83 MW Outlet backpressure 95 bar Blockage / valve latent GP / Modelica sigma 5.9 bar Pipe / Vflow 165 bar PCS state standby / 0.960 HIPPS state standby / 0.936 Demand Pmax 253 ± 8.1 bar Capacity Pcap 239 ± 10 bar Burst 2.60e-3

The PDT combines process demand, barrier reliability and structural capacity into burst probability.

3. Pipe Structural Reliability

Demand versus capacity

R = 0.143

Failure occurs when demand exceeds capacity: Pf = P(Pmax >= Pcap).

4. Risk State

Warning: risk above target

Decision support
Recommended action Monitor and prepare PCS/HIPPS intervention.
Current evidence path Physical sensors → digital twin → GP surrogate → PDT risk layer.

5. Real-time Forecast

Risk, pressure, flow and HIPPS reliability

Live
Burst pof Pressure Flow HIPPS reliability

Scenario Inputs

Prediction assumptions

These are what-if inputs for prediction, not direct manual operation of field equipment.

Presentation Script

How to explain the app