Editor's note: This is the third article in a five-article series on central-chiller-plant modeling. Part 1, "Primary/Secondary vs. Primary-Only Pumping", appeared in the April 2011 issue of HPAC Engineering. Part 2, "Efficient Control of a Primary/Secondary Plant", appeared in the May 2011 issue of HPAC Engineering.

The second article in this series arrived at primary/secondary- (P/S-) plant performance better than primary-only- (P-only-) plant performance when a P-only plant is operated with the minimum number of chillers/towers required to meet load. This article will develop a rationale for controlling a P-only plant at a level of performance on par with that of a P/S plant operating at the low-load delta-T conditions defined by the previous two articles.

P-Only-Plant Performance and Chiller Loading
The top chart in Figure 1 compares P/S- and P-only-plant performance, with the P/S plant having an advantage at site-load conditions of about 1,200 tons or less. The bottom chart shows chiller load in a P-only plant as load, wet-bulb temperature, and the number of chillers/towers in operation decreases. At a site load of 1,272 tons, the chiller is 49.5-percent loaded, suggesting the load might be met with one chiller/tower. One chiller/tower, however, cannot meet the load because when one chiller is on, the temperature of the water leaving the tower is higher, significantly increasing chiller lift.

P-Only-Plant Evaporator Flow and Velocity
In the P/S plant, flow in the evaporator is a constant 2,400 gpm; in the P-only plant, it varies, exceeding 2,400 gpm for virtually all site loads, as shown in the top chart in Figure 2. At 1,855 tons of site load, evaporator flow is approximately 3,500 gpm; at loads of 1,012 and 946 tons, it is more than 5,000 gpm.

The bottom chart in Figure 2 illustrates evaporator velocity. Evaporator velocity exceeds 10 fps at several site loads; at site loads of 1,012 and 946 tons, it exceeds 17 fps. Clearly, evaporator velocity must be controlled, if velocities of 10 to 12 fps are considered the maximum permissible, and control requires turning on another chiller. The design evaporator velocity for this chiller is 8.05 fps. The site loads were selected randomly; thus, there are other conditions at which evaporator velocity exceeds 12 fps. Generally, evaporator velocity reaches a maximum value at chiller loading of 90 percent or more and as a function of load delta-T.

System at 1,855 Tons
Figures 3 and 4 illustrate the modeled plant at energy equilibrium with two and three chillers/towers operating, respectively. Turning on a third chiller/tower reduced the load on each chiller from 82 percent to 51 percent and improved both chiller and plant performance. Plant power dropped from 1,271 to 1,249 kw, while evaporator velocity dropped from about 11.95 to 7.93 fps.

Page 2 of 2

Evaporator-Velocity Control
The top chart in Figure 5 illustrates P-only-plant performance with the minimum number of chillers on and controlled to maintain acceptable evaporator velocities. Another chiller was turned on for three conditions of site load. Although at 1,272 and 1,012 tons, one chiller would have met the load, the addition of a second chiller improved performance to approximately that of the P/S plant.

Comparing P-Only and P/S Plants
The top chart in Figure 6 compares the total pumping and tower-fan power of the two plants. The values are close, with the P/S plant using a little less power at several site loads. The top chart also gives the temperature of water entering the condenser, illustrating the effect of turning on another chiller in the P-only plant. The bottom chart in Figure 6 shows the chiller performance of the P-only plant generally is better. At 1,012 and 946 tons, the P/S values are high because the plant must provide about 39°F water from the evaporator to mix with bypass water (Figure 7).

Plant Performance and Number of Chillers/Towers Operating
Figure 8 shows the performance of the two plants is close when they are controlled as discussed above. Consider that, as shown in the bottom chart of Figure 8, for three site loads, the P-only plant requires more chillers/towers to be in operation. Also, note that the performance of the P/S chiller, as shown in the bottom chart of Figure 6, is not nearly as good as that of the P-only chiller at 1,012 and 946 site tons, although the performance of the two plants is about the same.

Figure 9 shows the performance of the four plants studied in this series thus far. The suggested control strategies significantly improve plant performance, whether the plant utilizes P/S or P-only distribution pumping.

Summary
Figure 1 shows the P/S plant with temperature-of-water-entering-the-air-handler-coil control performs better than the P-only plant with the minimum number of chillers/towers operating. Figure 2 illustrates the reason: Evaporator flow and velocity are significantly greater than design and, therefore, result in high secondary-pump power in the P-only plant.

Figures 3 and 4 provide the parameters of the P-only plant operating at a 1,855-ton site load with two and three chillers/towers, respectively. The need to control evaporator velocity is established. Figure 5 shows the improvement in plant performance resulting from turning on another chiller and, thus, reducing evaporator velocity and secondary-pump power.

Figure 6 introduces the P/S plant with control of the temperature of water entering the air-handler coil and compares the two plants’ chiller performance, showing that the increased number of chillers in operation in the P-only plant improves performance. Figure 8 shows the performance of the two plants is about the same, if they are controlled as described above.

Conclusion
The first article in this series established that bypass control is a bad strategy for a P/S plant. The first article also established the performance of a P-only plant with the minimum number of chillers on. The P-only plant performed better, but both plants could have been controlled more efficiently.

The second article evaluated the P/S plant and the concept of controlling the temperature of water entering the air-handler coil. The result was a significant improvement in the performance of the P/S plant, as seen in Figure 9.

This article discussed improvement of P-only-plant performance and identified the problem of high evaporator velocity. Turning on another chiller/tower resulted in improved plant performance, as shown in Figure 8. The next article will discuss low-load delta-T, suggesting it is inevitable if a secondary pump is controlled with a differential-pressure transmitter.

Kirby Nelson, PE, has been involved in the modeling of HVAC systems since the oil embargo of 1973—first as corporate energy manager for Texas Instruments Inc., then as a consultant. Models he has used include DOE-2, E Cube, and models developed on an analog/digital computer, including models of cleanrooms. A life member of the American Society of Heating, Refrigerating and Air-Conditioning Engineers, he has presented numerous papers, led an energy engineering delegation to China, and more recently developed models for district cooling systems, thermal-storage systems, and central plants.