Coarse modeling and simulation of a hydraulic pipe network
Assessing the availability of utilities like cooling water is an important task in brownfield projects. The suitability of existing infrastructure includes assessment of the existing pipe network with the additional flow rates for the new equipment. This task can become increasingly challenging if the equipment share utilities in a large site with several production units and equipment. The concept discussed in this post can be used for modeling and simulating a pipe network, in absence of a comprehensive simulation model.
Key Idea: Instead of detailing each equipment in different production units, each production unit is represented by pressure drop device such as control valve.
Objective: This exercise of coarse modeling is useful to save time and resources during preliminary analysis of the pipe network for bottlenecks, evaluating extension of network with additional equipment and trouble-shooting equipment issues.
Modeling & Simulation Procedure: Let us consider a case study, based on one of my industrial projects. The software tool used for modeling in this exercise was Pipe-flow. The software has easy to use help, so I'll skip the introduction available with the software.
Consider the plant layout shown in Figure 1, PT-1 to PT-4 are production units on the site with several heat exchange equipment and EQ-1 is a standalone compressor equipment. CT5 is a cooling tower which supply water to the equipment on site. The EQ-New is the new equipment which has to be supplied cooling water from CT5. The modeling procedure for this task can be divided into the four steps.
1. System identification
The main task here is collecting key information; pipe length, diameter, elevation, valve types, joints, bends, material of construction and cooling tower pump capacity. This is useful for building the model using software tools. The fluid properties are based on water at operating temperature.
At this step, it is also useful to identify the location of flow and pressure sensor for verifying process model later. The location of these sensors is also useful to decide the which equipment can be lumped together as one. Useful documents for identifying the system are plot plan, PIDs, isometric drawings and pump curves.
Tip: Verify documents with installation in field.
Key Idea: Instead of detailing each equipment in different production units, each production unit is represented by pressure drop device such as control valve.
Objective: This exercise of coarse modeling is useful to save time and resources during preliminary analysis of the pipe network for bottlenecks, evaluating extension of network with additional equipment and trouble-shooting equipment issues.
Modeling & Simulation Procedure: Let us consider a case study, based on one of my industrial projects. The software tool used for modeling in this exercise was Pipe-flow. The software has easy to use help, so I'll skip the introduction available with the software.
Consider the plant layout shown in Figure 1, PT-1 to PT-4 are production units on the site with several heat exchange equipment and EQ-1 is a standalone compressor equipment. CT5 is a cooling tower which supply water to the equipment on site. The EQ-New is the new equipment which has to be supplied cooling water from CT5. The modeling procedure for this task can be divided into the four steps.
 |
| Fig1: Site Layout |
The main task here is collecting key information; pipe length, diameter, elevation, valve types, joints, bends, material of construction and cooling tower pump capacity. This is useful for building the model using software tools. The fluid properties are based on water at operating temperature.
At this step, it is also useful to identify the location of flow and pressure sensor for verifying process model later. The location of these sensors is also useful to decide the which equipment can be lumped together as one. Useful documents for identifying the system are plot plan, PIDs, isometric drawings and pump curves.
Tip: Verify documents with installation in field.
 |
| Fig2: System Identification |
The next step is to translating the information to mathematical model using the software (Pipe-Flow). The network can be drawn using the toolbox and then selecting basic information regarding pipe size, length, bends, valves, etc. The cooling tower and pumps (with performance curves) are can also be built in the model.
Selecting the friction model is critical. The pressure drop is calculated based on Darcy-Weisbach relations, since Hagen-Poisule law is valid only for laminar flow and usually in industrial applications fluid flows are turbulent. Flow regime can also be checked in simulations.
To lump each group of equipment (or production plants on site) as a single pressure drop device, preferably use a control valve (sizing valve) for each group. Measured flow rate through the plant is used to define the valve, and pressure drop is calculated by the software. Theoretically, it is also possible to use 'flow dependent' or 'fixed' pressure drop device. However, this would require knowledge of pressure drop across the plants, which is usually not measured.
3. Model verification
To lump each group of equipment (or production plants on site) as a single pressure drop device, preferably use a control valve (sizing valve) for each group. Measured flow rate through the plant is used to define the valve, and pressure drop is calculated by the software. Theoretically, it is also possible to use 'flow dependent' or 'fixed' pressure drop device. However, this would require knowledge of pressure drop across the plants, which is usually not measured.
 |
| Fig3: Model Building with control valves |
The model can be verified based on normal operation and production plant maintenance cases. The maintenance cases can be shutdown of one or more equipment on the site. The flow and pressure from field measurement are needed to verify the model against calculated flow and pressure. Field log and online log can be useful for this information. If available and possible, flow can be measured with ultrasonic flow measurement devices and temporary pressure gauges.
Note: While using control valves for representing the plant, flow rate through the plant is fixed. Thus the calculate pressure is verified with the field measurements.
Note: While using control valves for representing the plant, flow rate through the plant is fixed. Thus the calculate pressure is verified with the field measurements.
 |
| Fig4: Model verification against measured flow and pressure |
Now this is the main step of the procedure, here you build and simulate scenarios based on the objective. In this task, the network was evaluated with the new equipment and cooling water pumps (in red).
To draw valid conclusions from the simulation it is important to remember the assumptions. By assuming the plant as a control valve (sizing valve) the flow rate is fixed. In the real plant, since lumped equipment doesn’t change, decrease in pressure drop across a plant is not feasible without the decreased flow. Hence, changes in pressure drop across the lumped equipment needs to be observed and compared with the base case. There are three possible changes in calculated pressure drop across the representative control valves:
 |
| Fig5: Simulating Scenarios |
- The pressure drop remains same for the same required flow rate: This means the change in the network would not affect the cooling water flow rate through the plant.
- The pressure drop increases for the same required flow rate: Thus the cooling water discharge from the equipment in the plant may need to be throttled so that other units in the network receive sufficient cooling water.
- The pressure drop decreases for the same required flow rate: This indicates need for cooling water booster pump for the plant to achieve the required flow.
Tip: Booster pumps in individual plants is useful to overcome plant pressure drop or elevation. The overall supply of cooling water is limited by the main cooling water sump pump.
Further steps to debottleneck the network would include detailing parts of the network (production plants and equipment) where the pressure drop decreases, to identify the location and size of booster pumps.
Underlying Concept: A useful analogy for analyzing the cooling water system is the treatment of hydraulic network as an electrical circuit. The flow resistant can be added in series or parallel to understand the impact on the network. The booster pumps in the network would reduce the pressure drop (flow resistance) for the plant in the network.
The system shown in Fig2 can be simplified and analyzed as shown below.
Further steps to debottleneck the network would include detailing parts of the network (production plants and equipment) where the pressure drop decreases, to identify the location and size of booster pumps.
Underlying Concept: A useful analogy for analyzing the cooling water system is the treatment of hydraulic network as an electrical circuit. The flow resistant can be added in series or parallel to understand the impact on the network. The booster pumps in the network would reduce the pressure drop (flow resistance) for the plant in the network.
 |
| Fig6: Important relations for analogy |
 |
| Fig7: Pipe network as hydraulic circuit |
