ASHRAE Journal - July 2010 - 39

low pump operating frequency. An extensive simulation study also was conducted to further compare the energy performances of both strategies with normal cooling loads in the zone (fully occupied) and evaluate their annual energy performance (which cannot be tested on site).
Site Tests

Mset

Mup

The energy performance of both strategies was tested on site. During the tests, the temperature setpoint at the secondary side, chiller supplied water temperature setpoint and the operating numbers of all related components (e.g., heat exchangers and water pumps) were set the same for both strategies. Taking into account the light load in the zone at present, the pressure differential setpoint in the original strategy was set at 20 kPa (2.9 psi) instead of at the design value of 60 kPa (8.7 psi). When using the original strategy, the power consumption of the pump was 14.1 kW. When using the alternative strategy, the actual pressure drop across the heat exchanger loop was 8.3 kPa (1.2 psi) and the power consumption of the pump was reduced to 11.0 kW. The energy saving due to the use of the alternative control strategy was 22%. This demonstrated that the alternative strategy has better energy performance than that of the original. The site application of the alternative control strategy over several months also demonstrated that this strategy can provide stable and reliable control in practical applications.
Simulation Tests

Mlow

100%

PI Output

Figure 4: Relationship between the PI temperature output and the water flow setpoint.

The simulation tests used a virtual building system that was constructed for previous studies.7,8 Detailed operational performance of the variable speed pumps distributing water to the heat exchangers of only Zone 1 is presented in this article as the example. The cooling loads used in the tests were predetermined using EnergyPlus9 based on the design data and hourly based weather data of the typical year in Hong Kong. Since the operation of variable speed pumps distributing water to heat exchangers has an impact on the operation of the pumps on the secondary side and chillers, the control strategies used for these components are introduced here briefly. The chillers were sequenced based on their design cooling capacities due to one constant primary chilled water pump and one constant condenser water pump dedicated to one chiller in this system. A threshold of 10% of the design cooling capacity was used for bringing chillers online and offline. For variable speed pumps and heat exchangers, they were sequenced based on 85% of their design water flow rates. A threshold of 10% of the design flow rates was used for switching the pumps and heat exchangers on and off. For stable control, a minimal time interval was introduced in these sequence strategies to avoid frequent switching. The speed of pumps distributing water to terminal units were controlled through resetting the pressure differential setpoint at the critical loop while the setpoint was optimized using the optimization strategy presented in ASHRAE Handbook,10 in which the water valve positions of terminal units were used to determine the increase or decrease of the pressure differential
July 2010

setpoint based on the last setting, by a fixed increment.10 Since the chilled water supply temperature setpoints at the chiller side and secondary side of heat exchangers have significant impacts on the energy consumptions of chillers and pumps, the design setpoints were used and remained unchanged during the tests. The operational data of the tests during three typical working days using the previous control strategies are presented to compare the energy performances of both strategies. These typical working days represent the typical operating conditions of the air-conditioning system in the spring, mild summer and sunny summer. As shown in Table 1, compared with the original strategy, the alternative strategy saved 182.3 kWh (16.26%), 202.9 kWh (15.28%) and 186.2 kWh (11.25%) of energy of the primary pumps in the three typical days. Table 1 shows that when using the alternative strategy, the energy consumption of the secondary pumps was reduced slightly while the energy consumption of the chillers was increased slightly in the three test days. The pump energy reduction is attributed to, at very light load conditions, a lower (i.e., lower than its setpoint) supply water temperature at the secondary side of heat exchangers was provided by the alternative strategy. This was due to the introduction of the lower limit of 20 Hz for the pump operating frequency. This lower water supply temperature resulted in saving some energy of the secondary pumps. The increase of the chiller energy is probably due to the increase of the latent load resulted by the lower supply water temperature to terminal units, or the effects of system strong dynamics in the simulation, etc. During the tests, the pressure differential setpoint used in the original control strategy was the design value of 60 kPa (8.7 psi). The monthly energy consumption of the primary and secondary pumps in Zone 1 and the total monthly energy consumption of the chilled water system (including the energy consumptions of chillers, all variable speed pumps in Zone 1 and constant speed pumps dedicated with chillers) using both strategies was simulated, and the results are presented in Table 2.
ASHRAE Journal 39

ASHRAE Journal - July 2010

Table of Contents for the Digital Edition of ASHRAE Journal - July 2010