Safety is an emergent property of a system not a component property. So how do we ensure a holistic view is taken during design? We believe the answer partly lies in the use of virtual system simulation (VSS) starting at requirements and hazard analysis, and then continuously through to commissioning and operations. Berkeley & Imperial are amongst the best placed to support your organization's transition to a fully cyber-physical operation.
These VSS can be used to support conceptual design, safety engineering, design verification, fault-finding, and validation testing. VSS models allow Berkeley & Imperial to evaluate the performance of each functional block of the system and the modular nature of each VSS facilitates plug & play for hardware-in-the-loop testing (HIL). Fault injection into the VSS can be used to shake down the system repeatedly testing what happens when a component fails.
Berkeley & Imperial uses state-of-the-art simulation engines, and in-house tools and expertise, to create virtual models of client systems. The challenge is to build VSS models that represent the multi-physics of real and complex systems, and to use the same models during software testing, hardware testing, training, etc. The end result of the MBE workflow is a model that can be used as a digital twin supporting operations, troubleshooting and maintenance.
Berkeley & Imperial has tied together 3D CAD models with VSS models of hydraulic modules created with Modelica language. These models have then been controlled through Wolfram language, thereby creating a virtual system simulator.
The modeled industrial unit is used to provide hydraulic power for mechanical handling. The model includes virtualized control, power, and hydraulics modules, driving hydraulic fluid movement in forward and reverse modes.
Berkeley & Imperial were asked to create a virtual system simulation (VSS) for a hydraulic control unit. The model includes hydraulics, mechatronics, and automation. The resulting model required the solution of 2619 differential algebraic equations. A key challenge was to ensure the model would run 'real-time' to support Hardware-In-the-Loop (HIL) testing.
State-of-the art tools supporting virtualization, software-in-the-loop testing, and hardware-in-the-loop testing also provide a platform for training and competency assessment. Berkeley & Imperial has combined these state-of-the-art tools with web-based learning tools in order to create an online, modular Training Academy targeted at Conventional and Managed Pressure Drilling. The academy includes student registrartion, structured learning paths, and both intermediate and final exams.
Berkeley & Imperial has carried out reliability, probability, and stochastic analysis for a wide range of applications. Over the years we have used various methods that include event trees, fault trees, reliability block diagrams, and Markov state transition. Each method has its weaknesses, which in one particular case led us to develop our own software so that we could model the effects of correlation, complex interaction, voting, and more.
Berkeley & Imperial has developed a set of tools for modeling the dispersion of atmospheric ‘pollutants’. These tools include a monte carlo simulator that can call on either our Gaussian dispersion model, or our implementation of the SLAB heavy gas dispersion model. Both use Pasquill atmospheric stability classfications and Briggs dispersion coefficients.
Our Gaussian dispersion model includes the effects of aerosols, and atmospheric boundary layer inversions. Projects supported by Berkeley & Imperial include decision-support for a major CO2 production company who encountered high levels of radon gas during field appraisal.
Berkeley & Imperial uses a set of tools for modeling gas, liquid, and two-phase fluids. These models include characterization of mass flow rate and pressure at the exit plane and fully expanded jet.
The tools include both phenomenological models like ATHENS, and OpenFoam CFD models. Output results can be overlaid on platform and plant layout drawings to aid conmmunication of the impact of hazard effects and consequences.
Berkeley & Imperial has developed unique insights on how complexity hinders the automation of drilling operations. Supplementing the human with automated responses requires complete and sufficient understanding of the large number of operating states, and the control and mitigation of faults. This is an enormously complex problem but can be done. Efforts to enumerate the states of a system should extend existing engineering methods and tools, so we developed a process for enumerating the Piping & Instrumentation Diagram (P&ID). This allows us to record an initial state, and then progressively change the system state. The whole process is visual, point and click, and provides a means for creating the state matrix for every valve, pump, isolable selection, etc. It also captures pressurization state, sequence, timing, interlocks and alarms.
Every now and then Berkeley & Imperial is fortunate, and we have the chance to work on something challenging and fascinating. One example was helping Shell’s Game-Changer team develop software that combined optimum well-path alignment, with an artificial, impressed, geomagnetic field. This work called on our deep Mathematica skills, and led us into research on vector algebra for Euclidean geometry, and quaternions. The results from this analysis were used to drill shallow, horizontal and challenging wells.