Closed Loop Systems: Known Pressure and Reservoir Boundary Nodes

A closed-loop system may be fully connected yet remain hydraulically undefined. Even with every pipe, fitting, heat exchanger, and pump represented, the model may still be unsolvable if it lacks a pressure reference.

The pressure-reference boundary must represent the real system, not merely provide a convenient way to complete the flowsheet. For systems with a defined pressure-setting interface, FluidFlow provides two relevant boundary representations: Known Pressure or Reservoir node. Choosing between them depends on connectivity, liquid-level effects, connection elevations, and the physical behavior of the pressure interface.

In a closed-loop arrangement, the working fluid circulates through a connected path and returns through the system. A booster supplies the pressure rise needed to overcome pipe friction, fitting losses, valve resistance, heat-exchanger pressure drop, and other hydraulic resistance.

The pressure rise and resistance terms establish pressure changes around the circuit. They do not independently establish the absolute pressure level throughout the system. The model still needs a defined pressure condition from which the pressure profile can be calculated.

FluidFlow applies conservation of mass and energy across the connected network. Its energy balance accounts for pressure, velocity, elevation, energy added by boosters, and energy dissipated through resistance. To establish the pressure profile, the model requires a specified pressure at at least one boundary.

This is why adding a pump alone is not sufficient. A pump provides differential pressure. It does not, by itself, define the absolute operating pressure at the pump suction, high points, equipment inlets, or other locations in the loop.

The boundary represents the physical interface that establishes the system pressure condition. Depending on the installation, that interface may be an expansion vessel, accumulator, pressurized header, or another part of the plant that is outside the modeled network.

The selected node must reproduce the behavior of that interface. If the physical system has one connection at a specified pressure, a Known Pressure node may be appropriate. If the pressure reference boundary has multiple connections, a liquid level, or nozzle elevations that affect the available head, a Reservoir node provides a more complete representation.

Both components are entry or exit network points. FluidFlow calculates flow into or out of them according to mass and pressure balance. The engineer must therefore confirm that the selected boundary and its calculated flow direction are consistent with the actual system.

A Known Pressure node represents an entry or exit point where pressure is known or is a design condition. Only one pipe can connect to the node, and FluidFlow calculates the flow into or out of it.

Use a Known Pressure node when a single connection adequately represents the pressure-setting interface. Typical representations include a pressurized header, a vessel connection, a tank connection, or another defined location where the network exchanges fluid with a system outside the model.

The specified pressure must be applied at the correct elevation and using the appropriate pressure model. FluidFlow allows the boundary pressure to be defined as static or stagnation pressure.

Stagnation pressure is generally used to represent vessels, tanks, and large reservoirs, where velocity effects are negligible. Static pressure can represent a boundary in a pipe, such as a header or a pipeline tie-in, where the pressure condition corresponds to a pressure measured at the pipe wall, such as a pressure-gauge reading.

The main inputs are:

Figure 1. FluidFlow Input Editor for a Known Pressure node.

When a Known Pressure node represents a vessel or storage tank, liquid head can be included through the elevation or pressure definition. The selected basis must be applied consistently so liquid height is not omitted or counted twice.

A Known Pressure node should not be added solely because the model needs a pressure value. It must correspond to a real single-connection pressure interface. Otherwise, the model may solve while representing a source or sink that does not exist in the system.

A Reservoir node represents an entry or exit point where pressure is known or specified as a design condition. It is similar to a Known Pressure boundary but supports up to eight pipe connections.

Flow direction can differ between the connected branches. FluidFlow calculates flow into or out of the Reservoir through each branch according to the network mass and pressure balance.

Use a Reservoir when the physical pressure interface is a tank, vessel, or accumulator whose liquid level, surface pressure, or multiple nozzle connections affect the hydraulic model. This representation is particularly useful when several parts of the network connect to the same vessel at different elevations.

The principal inputs include:

Figure 2. FluidFlow Input Editor for a Reservoir node.

A Reservoir takes a single surface pressure input. FluidFlow treats that value as both the static and the stagnation pressure because velocity in large reservoirs is negligible compared with velocity in a pipe. The static-versus-stagnation pressure selection therefore applies to a Known Pressure node only.

The pressure available at a branch depends on the surface pressure, liquid head, and branch elevation. If the liquid level is below a branch connection, that connection may be located in the designated vapor space rather than the liquid region. For a vapor-space connection, the solver uses air instead of the node’s specified fluid. The liquid level and nozzle elevations must therefore match the physical vessel arrangement.

A Reservoir has no associated internal pressure drop in FluidFlow. Pressure changes assigned to its connected branches come from the defined vessel conditions, connection geometry, and any selected entrance or exit losses.

Figure 3. Setting entrance/exit losses via the Connections field of a Reservoir in the Input Editor.

In a closed loop, the circuit returns to its starting point, so pressure, velocity, and elevation at the start and end are identical. FluidFlow offers two modeling approaches.

Place a Reservoir in the main loop upstream of the pump, connect the pump downstream, and return the circuit to the Reservoir. The Reservoir defines the loop’s fluid, pressure, and temperature.

Figure 4. Closed loop modeled with an in-line Reservoir. The Reservoir (water, 15 °C, 1 atm surface pressure, 2 m liquid level above a vessel base at 0 m) sits upstream of the pump, and the circuit returns to it. Calculated pump duty: 22.61 m³/h at a 2.536 bar pressure rise.

Cut the loop at the reservoir location and replace the Reservoir with two Known Pressure nodes. Set both nodes to the same pressure and elevation; each node connects to one pipe, leaving an open path between them.

Figure 5. The same loop split at the Reservoir location, with two Known Pressure nodes in place of the Reservoir. Both nodes use stagnation pressure of 2 m Fluid g at an elevation of 0 m. The fluid (water) and temperature (15 °C) are defined at the inlet node. Calculated pump duty matches Figure 3: 22.61 m³/h at a 2.536 bar pressure rise.

For either setup, keep the boundary nodes on the main loop and define the fluid and temperature at the inlet boundary. Where practical, compare both representations as a model check. For step-by-step instructions, see Modeling closed-loop systems in FluidFlow.

The decision is not based on which symbol looks more like the P&ID. It is based on the hydraulic information that must be represented and the number and location of the real connections.

Start with the physical pressure interface and work outward.

This sequence prevents the boundary node from becoming an arbitrary pressure anchor disconnected from the actual system.

A manual loop calculation often focuses on the booster pressure rise required to overcome total friction at a target flow. That approach may establish differential pressure while leaving the absolute pressure profile unresolved.

The simplification becomes unreliable when elevation changes are significant, equipment inlet pressure is important, fluid properties vary with pressure, or a pressure-setting vessel connects to the system at a defined level and elevation. One assumed loop pressure can conceal local high- and low-pressure points.

A second failure occurs when vessel geometry is simplified too aggressively. Replacing a multi-connection tank with one Known Pressure node removes the individual branch heights and connection behavior. Conversely, using a Reservoir where the real interface is a single known-pressure pipe connection adds vessel behavior that the system may not have.

A third failure is accepting convergence as proof of correctness. A boundary can provide enough information for the equations to solve while still applying the wrong pressure model, elevation, liquid head, or flow direction.

The FluidFlow Liquid Module calculates pressure and temperature profiles, flow distribution, fluid properties, and equipment duties in open and closed-loop networks. It solves conservation of mass and energy across the connected system using the defined pipes, equipment, boosters, resistance, elevations, and boundaries.

A correctly defined Known Pressure or Reservoir boundary gives the model the pressure condition needed to calculate the connected pressure profile. The engineer can then examine pressures at pumps, heat exchangers, control valves, high points, low points, and branch junctions rather than relying on one assumed loop pressure.

The model also calculates flow into or out of the selected pressure boundary. That result is an essential validation check. The engineer must confirm that its direction and magnitude agree with the physical function of the interface and the operating case being modeled.

The model does not determine which plant item establishes system pressure. That must be confirmed from the P&ID, vessel data, operating description, equipment arrangement, and the actual connection to any accumulator, expansion arrangement, makeup system, or external header.

It also does not establish the correct operating pressure without an approved project basis. The engineer must define surface pressure, static or stagnation basis, reference elevation, liquid level, nozzle heights, and applicable operating cases.

A converged result does not prove that the pressure boundary is physically correct. Boundary meaning, calculated flow direction, liquid phase assignment, and the full system pressure profile must be reviewed together.

Yes, when it represents a real single-connection pressure interface through which fluid may enter or leave the modeled network. The specified pressure, elevation, and pressure model must correspond to the physical location and the applicable operating case. A closed loop can be modeled by splitting the model and using two Known Pressure nodes. Set both nodes to the same pressure and elevation. This approach produces the same results as the in-line Reservoir method when the pressure and reference elevations are identical.

Use a Reservoir when the pressure-setting interface is a tank, vessel, or accumulator with multiple connected pipes, a defined liquid level, or branch connection elevations that affect pressure. Use Known Pressure when one connection and a specified pressure condition adequately represent the interface.

Liquid level contributes to the hydrostatic head at each connected branch. If a branch is above the specified liquid level, it may connect to the vessel’s vapor space rather than the liquid region. In that case, the solver uses air rather than the node’s specified fluid. Incorrect liquid level or branch elevation can therefore affect the assigned fluid, phase, and pressure condition.

Use the basis that matches the physical design condition. For a Known Pressure node, stagnation pressure is used for vessels, tanks, and large reservoirs, while static pressure is used when the boundary represents a pipe or header location and the specified value corresponds to a pressure-gauge reading. A Reservoir takes a single surface pressure that serves as both.

Recalculate and review boundary flow direction, the complete pressure profile, local static and stagnation pressures, booster duty, elevations, liquid level, and the lowest-pressure locations. For a Reservoir, also verify each branch height and whether it connects to the intended liquid or vapor region. Confirm that the net flow across the Reservoir—the difference between total incoming and outgoing branch flows—is consistent with the modeled case, since the Reservoir can add or remove flow to maintain network mass balance.

Define the pressure-setting interface before finalizing the network topology. Establish whether the physical interface is a single known-pressure connection or a vessel whose level and multiple connections must be represented.

Where practical, compare both representations. Use Known Pressure for the first case and Reservoir for the second. Then validate the model through the calculated boundary flows and connected pressure profile rather than accepting convergence alone as evidence that the closed-loop system is represented correctly.