Hydraulic Digital Twin Verification of a Hybrid Boiler–Seawater Water-Loop Heat Pump System

Keywords: hydraulic digital twin, hydronic system verification, water-loop heat pump systems

The objective of this study is to emphasize the importance of verifying building mechanical system designs prior to construction. By evaluating system behaviour in a virtual environment under different operating scenarios, designers can optimize hydraulic performance and improve both capital investment (CAPEX) and long-term operating efficiency (OPEX). Problems identified only after installation are often difficult and costly to correct due to construction constraints.

The transition toward low-carbon building operation requires HVAC systems that move beyond conventional boiler–chiller configurations and enable more efficient energy redistribution within buildings. In this context, water-loop heat pump (WLHP) systems have gained increasing attention, particularly in buildings where heating and cooling demands occur simultaneously.

WLHP systems enable heat rejected from cooling-dominated zones to be reused in zones requiring heating, allowing internal heat recovery, and reducing auxiliary energy demand. However, the thermodynamic potential of such systems does not automatically translate into operational efficiency. System performance depends largely on the hydraulic behaviour of the common loop, including flow distribution, pressure stability, and part-load interaction between zones. Without proper hydraulic balance and pressure management, even high-efficiency heat pumps may fail to achieve the expected seasonal performance.

Therefore, the hydraulic design of WLHP systems requires careful coordination between piping configuration, pump operation, and control strategies. In complex multi-zone systems, conventional hydraulic calculations alone are often insufficient to reliably predict system behaviour.

To address this challenge, digital hydraulic modelling—often referred to as a hydraulic digital twin—can be used to simulate the interaction between pumps, control valves, heat pumps, and other hydraulic components under different operating conditions. This approach allows designers to evaluate system behaviour prior to construction and reduce the risk of operational problems during commissioning.

Case Study and System Configuration

Figure 1. The case study consists of two buildings located at the Bahçeşehir University Beşiktaş South Campus in Istanbul, Türkiye, approximately 30 m from the Bosphorus shoreline.

The primary academic building accommodates various educational functions including classrooms, auditoriums, library areas, and technical spaces, with a total enclosed area of approximately 25,000 m². A second building of approximately 2,600 m² houses administrative offices and support spaces.

A significant portion of the enclosed area is located below ground level, including a 480-seat conference hall, foyers, classrooms, library spaces, laboratories, and technical rooms. Due to the architectural configuration and internal heat gains from occupants, lighting, and equipment, simultaneous heating and cooling demands frequently occur within the facility.

The relatively mild winter climate of Istanbul and the coastal location of the campus further support the application of heat pump systems. Seawater temperatures at depths of 20–25 m remain relatively stable throughout the year (8–9°C in winter and approximately 20°C in summer), providing an effective heat rejection source for the cooling season while requiring auxiliary heating support during colder periods.

Considering these conditions, the HVAC system was designed as a hybrid boiler–seawater supported WLHP system. A central loop operating between 19–32°C serves as the building’s primary thermal balancing medium (Figure 2), with the piping network arranged according to the reverse-return principle to ensure hydraulic balance. The system operates with a 23–26°C dead-band to maximize internal heat recovery: heating is activated below 23°C via boiler-connected plate heat exchangers, while excess heat is rejected through seawater heat exchangers when the loop temperature exceeds 26°C. To increase thermal inertia, two 2,000 ℓ buffer tanks were integrated into the loop, extending the system’s residence time within the dead-band.

Figure 2. Overall hydraulic configuration of the WLHP system (Typical)

The hydronic distribution network is organized into multiple building zones served by ten secondary substations and two source stations (heating and cooling). Each zone supplies water-to-water heat pumps and water-cooled VRF units through secondary pumps. The system follows a primary–secondary pumping architecture, where zonal pump stations connect to the main loop via hydraulic decouplers to maintain approximately zero differential pressure (ΔP ≈ 0 kPa).

In summary, the auxiliary heating system consists of eight condensing boilers, each rated at 130 kW, producing 30°C low-temperature heating water and providing a total installed heating capacity of approximately 1,040 kW. The WLHP system serving the campus provides a total installed cooling capacity of approximately 3,103 kW.

Given the hydraulic complexity of this multi-zone system, verification of system behaviour became an important design step. A hydraulic digital twin model of the entire hydronic network was therefore developed prior to procurement and installation.

Hydraulic Digital Twin Modelling and System Verification

In complex HVAC hydronic systems, verifying whether the system will operate as intended after construction represents a major challenge during the design phase. While conventional hydraulic calculations are sufficient for pipe sizing and preliminary pump selection, they are often inadequate for predicting the behaviour of large multi-zone systems where multiple pumps, control valves, and terminal devices interact through a common distribution network.

To examine system behaviour under both peak-load and part-load conditions, a hydraulic digital twin of the entire hydronic network was created. The model was developed using Pipe-FLO Engineered software. Piping layout, lengths, diameters, fittings, and connection points were taken from the project’s Revit model, allowing the hydronic network to be represented with high geometric accuracy. Pump performance curves, valve characteristics, and pipe materials were also included in the model.

The primary objective of the modelling study was to verify the correct operation of the primary–secondary pumping arrangement. In this configuration, main loop pumps circulate water through the building loop, while secondary pumps supply water-to-water heat pumps and water-cooled VRF units. Hydraulic separation between the circuits is provided by hydraulic decouplers at each zone connection, working together with pressure-independent control valves (PICVs) on the return connections to regulate zonal flow and maintain the required differential pressure. The digital twin model was used to evaluate system behaviour under different load scenarios and confirm proper pump selection (Figure 3).

Figure 3. Prototype hydraulic digital twin model developed in Pipe-FLO.

The model enabled identification of the critical hydraulic circuit and determination of the differential pressure setpoint required to control the main loop pump speed, based on the most hydraulically remote control valve in the network. This approach ensures proper valve operation while avoiding excessive pump head and pumping energy. Additional simulations under typical part-load conditions confirmed stable pump operation, correct valve performance, and no unintended circulation through the hydraulic decouplers. Overall, the hydraulic digital twin verified system behaviour prior to procurement and installation, effectively serving as a form of virtual pre-commissioning during the design stage.

Conclusions

The performance of HVAC systems is strongly influenced by the hydraulic behaviour of the hydronic distribution network connecting pumps, control valves, and terminal devices. System energy performance therefore depends not only on the efficiency of individual equipment but also on the hydraulic performance of the distribution network. Hydraulic digital twin modeling provides an effective method for verifying complex hydronic systems during the design phase. By simulating different operating scenarios, designers can evaluate flow distribution, pressure conditions, pump operating ranges, and control valve performance before construction.
From an economic perspective, this approach benefits both investors and building operators by avoiding costly post-construction modifications and improving capital investment efficiency (CAPEX) as well as long-term operational performance (OPEX).