Printer Friendly

Dealing With Insufficient Plant Static Pressure.

Buildings exceeding the plant static pressure commonly experience negative pressure on large thermal loops, as design firm engineers and facility personnel may overlook its occurrence. A study was initialized by performing a pressure distribution analysis on the campus thermal loop to identify the reason for the negative pressure. Further investigation into several solutions was required, as the issue is not easily solved by increasing the building pump's capacity or balancing the water system.

Raising the system static pressure, indirect connection with a heat exchanger, and installing a pressure sustaining valve (PSV) were all explored thoroughly as separate solutions by data analysis and simulation. Raising the system static pressure and indirect connection proved to be costly solutions due to the need for expensive new double-capacity expansion tanks and installation of higher-pressure capacity equipment, or building renovations due to limited mechanical room space.

The PSV proved to be the most effective solution after an in-depth analysis. Simulations of the PSV displayed the reliability of its pressure control as well as any changes in power, head, and flow in the pumps. No increase in pumping power was observed with the PSV installation. The control scheme of the pump and setpoint of the PSV are dependent upon the height and demand of the building. Also, installing the PSV on the building return pipe will not change plant operation and pressure control. The solutions and procedures outlined provide a beneficial technical guide or lessons learned for large thermal loop and building design, operation, and maintenance.

The Texas A&M University (TAMU) Utilities and Energy Services (UES) has been working to increase efficiency and reduce energy consumption through many initiatives, including thermal loop and building optimization. Since its inception in 2002, TAMU energy conservation activities have saved $200 million and reduced energy consumption per square foot by 45%.

The main campus of Texas A&M University has two thermal plants, the central utility plant (CUP) and satellite utility plant 3 (SUP3), which provide chilled water (CHW) and heating hot water (HHW) through 98,000 ft (29 870 m) of pipe to all buildings on the main campus, as seen in Figure 1. The plants contain variable frequency electric chillers, boilers with turndown ratios, and primary CHW and HHW pumps, allowing adjusting of production and flow rate to meet demand. On the thermal loop, four of the 230 buildings are experiencing negative top coil pressure, two of which (Kyle Field and Eller O&M) will be the focus of this study along with simulations of plant pressure.

Commonly, buildings exceeding the plant static pressure level experience negative pressure in their tallest parts. To identify the reason for the negative pressure, a pressure distribution analysis was performed on the campus thermal loop. Once the negative pressure was verified, raising the system static pressure, indirect connection with a heat exchanger, and installing a PSV on return lines were all explored thoroughly as separate solutions by analyzing the pressure distribution and running simulations. The simulations of the selected PSV installation displayed the differences between having a PSV installed and no PSV installed, as well as any changes in power, head, and flow in the pumps. The solutions and procedures outlined will provide a beneficial technical guide for large thermal loop design, operation, and maintenance as a previous standard was not found.

Potential Solutions for the Negative Pressure Issue

Raise Plant Static Pressure

One way to resolve the negative pressure issue is to raise the plant static pressure. Two expansion tanks for the thermal network pressure control are in operation at Texas A&M, both located on the suction side of the plant pumps at SUP3. One of the primary functions of an expansion tank, aside from accounting for thermal expansion of the water, is to pressurize the system so water can reach the tallest building without negative pressure. This method makes sure thermal water fills all of the pipes and coils for each building, even the taller buildings. To partly alleviate the problem, the expansion tank pressure was raised from 30 psi (206.84 kPa) to 50 psi (344.74 kPa) in June 2015. An upper limit of 50 psi (344.74 kPa) was discovered due to existing pipe leakage; however, 50 psi (344.74 kPa) is still not high enough to satisfy all buildings on campus.

Raising the plant expansion tank pressure at SUP3 to an elevation of 353 ft (107.59 m), shown in Figure 2, is sufficient to satisfy the static pressure requirements of the tallest building on the loop. After raising the static pressure in the expansion tank, the pumps no longer have to overcome both frictional losses and pump water to the coils above static pressure. Rather, the pump's primary role in the system is just to make up for frictional losses as the expansion tank ensures water reaches the top coils. When the expansion tank does its job correctly, it will significantly reduce the workload of all the building pumps in the system and extend their operating life while reducing electricity consumption.

From these assumptions, the supply pressure is 117 psig (806.69 kPa), the plant pump pressure is 127 psig (875.64 kPa), and the plant return pressure is 97 psig (668.79 kPa). These pressures were developed to satisfy Eller O&M's 550 ft (167.64 m) elevation. To comply with this option's pressure profile, an expansion tank twice the size of the current one would be needed to be set at 97 psi (668.79 kPa), an increase from 50 psi (344.74 kPa), to counter thermal expansion and to cover the tallest building. This calculation was done using a diaphragm tank volume formula from Chapter 13 of the 2012 ASHRAE Handbook--HVAC Systems and Equipment. (1)

Increasing the supply pressure in the loop can lead to an increase of pipe leakage, especially in older pipes like those on the Texas A&M campus. Each boiler/chiller also has a relief valve. If the system pressure were to be raised, each relief valve's limit would have to be increased, if the valve would even be capable, but also the relief valve's limit is set to the boiler/chiller's existing capacity, so several boilers and chillers, as well as the expansion tank, may need to be replaced to account for a larger capacity. These facts suddenly make this option an expensive one to pursue, so others will be explored.

Indirect Connection

The typical way of dealing with this issue, especially in new buildings being constructed, is to install a heat exchanger to indirectly connect back into the campus's thermal loop. The heat exchanger will isolate the building's hydronic system from that of the campus's and allow water to reach the top AHU coils by using an isolated building pump system without negative pressure and will not interfere with the campus thermal loop static pressure.

Cain Hall, a new building under construction on Texas A&M's main campus, is being fitted for a heat exchanger. Its elevation is right at the static pressure level and, according to TAMU design standards, requires a heat exchanger for indirect connection to the campus thermal loop.

However, since Kyle Field recently underwent renovations, its mechanical rooms, as well as many of the other buildings experiencing this problem, are not fitted to allow an extra heat exchanger. Due to spacing and the need for more renovations, this option also becomes too expensive.

Pressure Sustaining Valve (PSV)

The objective of this option is to use the building pumps and PSV to regulate the pressure profile at each building that is taller than the system pressure setpoint without changing the plant configuration. The PSV works by modulating position to maintain an upstream pressure setpoint, and the PSV is operated through either a stand-alone controller or linking building control program. Each PSV controller and building pump controls are set to the building's height.

The campus hydronic system is made up of a supply line delivering water to each building, the building's hydronic system circulating the supplied water throughout, and then delivering the water to a return line that returns the water back to the central plant. Figure 3 is a simplified version of the hydronic loop. Building pumps are operated on differential pressure for each building, and plant pumps deliver required flow calculated from each building's automation system and totaled in the plant's control system.

Figure 3 represents a simplified diagram of the HHW/CHW system with a pump and PSV at the building side. The plant delivers water to the building, and the building pumps receive the water at a pressure of [] and then distribute it to the building pipes at a pressure of [P.sub.discharge]. The plant receives the return water back at a pressure of [P.sub.return]. Although the PSV setpoint value, [P.sub.setpoint], will be different for each building, the assumptions will be used that the PSV setpoint will be held constant and is customized based on the building to allow at least 5 psi (34.47 kPa) at the top.

If the building has a variable speed drive pump, the pumping speed will change as the pressure drop across the building changes with the flow rate. When the building side pump is shut off completely, the PSV will hold water in the return to maintain 5 psig (34.47 kPa) at the top of each building.

Simulation for PSV and Without PSV

Simulation Model and Calibration

The Eller O&M building is a 15-story 109,609 [ft.sup.2] (10 183 [m.sup.2]) building that opened in 1973 on the main campus of Texas A&M. The building rises above the static pressure provided by the plant. Kyle Field underwent renovations during 2013 to 2015, amounting to a $450 million project that raised the stadium's capacity to 102,733 from 82,600 people. Prior to the renovation, negative static pressure was not a concern at the stadium.

To analyze and evaluate the current thermal distribution system, providing inputs regarding impacts and potential changes to the current distribution system, a hydraulic simulation was developed. The simulation was developed to model the supply and return pressures, flows, and temperatures of the loop during peak times of demand. The model was calibrated using the actual measured flow rates and pressures and reliable meter data at CUP and SUP3. Then these values were matched in the hydronic model. This allowed a solid starting point for the introduction of the PSV into the system modeled to see what would happen. The key assumption in modeling is an incompressible fluid.

Results from the simulation showed negative pressures at four different points in the loop: Kyle Field, Rudder Tower, Eller O&M, and Richardson. These are the tallest buildings on campus, and the elevation input into the model showed that the height of the buildings was higher than the return pressure of the primary return loop. These negative pressures were then later confirmed through field measurements taken at the top pipes of the buildings. This model of the entire main campus helped identify problems in the loop, the most outstanding being the negative pressure at the top coils of these four buildings.

To further investigate this issue and understand the nature of each of the building's thermal piping and pumping system, specific models were made for Kyle Field, which developed negative pressure after undergoing renovations that increased the height of the structure, and Eller O&M, the building with the highest elevation. These individual models sought to answer several questions, including the effect of adding a pressure sustaining valve on the return pipes on the top coil pressure, on the pumping power, and on the building's control valves. The balance between the return pressure and the pumping capacity had to be understood to determine if the system specifications could handle a pressure sustaining valve.

Negative Pressure Study

The development of the models allows a detailed pressure study on the top coils of both Kyle Field and Eller O&M. The top coil elevation was adjusted to model the effect of elevation on the inlet and outlet pressure of the coil, before and after the flow control valve into the coil. Figure 4 shows the differential pressure between inlet and outlet of the top coil at 16 psi (110.32 kPa) for Eller O&M and 21 psi (144.79 kPa) for Kyle Field. This relationship determined the outlet pressure of the top coils. The figure allows an understanding of the relationship between the inlet and outlet pressure profiles--at a certain elevation, negative pressure will develop. The simulations were based on peak demand times. For Kyle Field and Eller O&M, this is a 57 psi (393 kPa) primary return pressure and 51 psi (351.63 kPa) primary supply pressure, from trended data.

The conclusion drawn from Figure 4 is that elevation directly relates to the negative outlet pressure on the top coil, and that the differential pressure across the coil and flow control valve also contribute, as it causes a drop in the inlet pressure. This can be seen in Eller O&M as negative pressure begins to develop around 180 ft (54.86 m) relative elevation. The solution to this issue is to raise the outlet pressure to an acceptable positive value while at the same time ensuring adequate flow through the valve. This should be accomplished while also limiting any increase in electrical consumption of the pumps.

Figure 5 shows, with varying primary return pressure, the allowable top coil elevation for both Eller O&M and Kyle Field to achieve zero coil pressure. The slopes of the lines are the same, but the upward shift of the Kyle Field line shows the effect pipe friction loss has on the top coil pressure. Looking from the return side with a pressure of 50 psi (344.74 kPa) and moving up the pipe to the top coil, pressure will decrease, but pressure lost due to pipe friction will be regained. Due to this phenomenon, there is a slight difference in the lines of zero pressure between Eller and Kyle Field because of differences in pipe friction. Figure 5 shows that as primary return pressure increases, the top coil outlet pressure also increases, giving credibility to the PSV solution--if the primary return pressure can be raised to a value corresponding to the elevation of the top coil, adequate pressure will result.

PSV Effect Study

The building CHW system with the pressure sustaining valve installed is modeled based on real data. The actual performance of the valve was entered as an input, so the modeling of the sustaining valves could be trusted. With this real performance data input, the valve's performance with the varying supply and return pressures, as well as flow rates, makes the entire model more reliable. Every building pump modeled, in both the Kyle Field and Eller O&M simulations, has the real pump curve input and its speed percentage, horsepower, discharge pressure, and flow rate are all built on actual performance.

First using the model of Kyle Field, several scenarios of the PSV were created. The goal was to test the validity of the PSV installation raising the negative pressure in the top coils as well as ensuring existing pumping capacity was sufficient. By simulating several scenarios of varying pressures and flow, with and without the PSV, the net positive or negative effect of the PSV could be documented and its performance reviewed.

PSV Effect on Primary Return. In Table 1, with varying return pressures, the average horsepower per pump is the same with the PSV as without. After careful investigation, it was shown that the PSV increased the back pressure on the flow control valves and that the flow control valves on the air-handling unit opened up to compensate for the decrease in differential pressure across the valve. It was also observed that pumping speed and head was also the same with and without the PSV, 72% and 38 psi (101 kPa), respectively, for Kyle Field and 95% and 47.6 psi (329 kPa), respectively, for Eller O&M.

A more thorough review of the results shows that the set discharge pressure of the pumps causes an adequate supply of pressure through the uppermost coils, such that increasing return pressure to a positive amount continues to allow sufficient flow through the control valves of the coils, allowing that adequate differential pressure exists across the valve.

Figure 6 attempts to graphically illustrate the steady, reliable nature of the PSV. It has been observed from trended field data that return pressure varies constantly, illustrated here by the range of return pressure. Due to this it would be difficult to ensure adequate pressure at the top coil, and the fluctuation in return pressure would cause unreliable control, as seen in the scenario without the PSV. This figure illustrates that with the installation of the PSV, no matter the primary return pressure, the top coil pressure is reliably positive and steady.

PSV Effect on Flow Rate. Increasing flow rate slightly decreases the inlet pressure for both Kyle Field and Eller O&M, as seen in Table 2, with a 1 psi (6.89 kPa) drop for Kyle Field and a 0.6 psi (4.14 kPa) drop in Eller O&M. Outlet pressure, on the other hand, has a slight increase in pressure as flow rate increases for all but the no PSV Eller O&M simulation, and is drastically different between the PSV case versus the no PSV case. As flow rate increases from 40% to 100%, pumping power consumption goes up by 4.3 hp (3.21 kW) per pump for Kyle Field and by 4.6 hp (3.43 kW) per pump for Eller O&M. This is a large increase in pumping power. However, pumping power consumption is not different for the PSV versus no PSV case.


Several options exist for solving the negative pressure issue. Possibilities were raising system pressure or indirect connection by a heat exchanger. Both proved to be costly pursuits given the current equipment and situations. The solution pursued was installing a PSV on return pipe at the buildings where the elevation is higher than that of the system static pressure and is a cost-effective option when compared to the above two options. The control scheme of the pump and setpoint of the pressure sustaining valve are dependent upon the height and demand of the building. Also, installing the PSV on the building return pipe will not change plant operation and pressure control.

The analysis and simulation provide a very useful solution of how to solve negative pressure for top cooling coils, improving large loop performance. As a practical, technical guide providing pumping operation and PSV analysis connected to district thermal systems, it puts forth a solution to problems of air introduction, corrosion, and ensuring that air-handling units are adequately provided with flow. This study led to Texas A&M University's new design standards for the interconnection of building hydronic systems to campus thermal utility infrastructure and has been implemented to resolve several of the problematic buildings.


(1.) 2012 ASHRAE Handbook--HVAC Systems and Equipment.

Hui Chen, P.E., is a professional engineer; James Riley, Les Williams, and Homer Bruner Jr., are executive director, director and assistant director; Amy Chen is energy engineer; and Robert Henry is a professional engineer at the Utilities and Energy Services, Texas A&M University.

Caption: FIGURE 1 Texas A&M main campus CHW (blue lines) and HHW (red lines).

Caption: FIGURE 2 Proposed increase in expansion tank pressure for the Eller O&M building.

Caption: FIGURE 3 PSV installation for the Eller O&M building.

Caption: FIGURE 4 Primary return pressure versus top coil outlet pressure for Kyle Field and Eller O&M.

Caption: FIGURE 5 Elevation versus primary return pressure at zero top coil pressure.

Caption: FIGURE 6 Varying primary return pressure versus top coil outlet pressure.
TABLE 1 Primary return pressure versus coil pressure pumping power
and speed.

         KYLE FIELD

           (PSI)      (PSI)      (PSI)   PER PUMP

PSV          50         22        7.6       10.8
             55         22        7.6       10.8
             60         22        7.6       10.8
             65         22        7.6       10.8
             70         22        7.7       10.8
No PSV       50         22      -12.0       10.8
             55         22       -7.2       10.8
             60         22       -2.5       10.8
             65         22        2.7       10.8
             70         22        2.7       10.8

         ELLER O&M

            (PSI)      (PSI)    PER PUMP

PSV          5.8        4.7         7.5
             5.8        4.7         7.5
             5.8        4.7         7.5
             5.8        4.7         7.5
             5.8        4.7         7.5
No PSV       5.8        -17         7.5
             5.8        -12         7.5
             5.8       -7.2         7.5
             5.8       -2.2         7.5
             5.8        2.8         7.5

TABLE 2 Flow rate versus coil pressure pumping horsepower
and speed.

        KYLE FIELD

         (%)      (PSI)      (PSI)   PER PUMP

PSV      40         23        5.0        6.5
         60         23        6.0        6.5
         80         22        7.0        7.0
        100         22        7.6       10.8
No       40         23       -7.5        6.5
PSV      60         23       -7.0        6.5
         80         22       -6.3        7.0
        100         22       -5.0       10.8

        ELLER O&M

           (PSI)      (PSI)   PER PUMP

PSV         6.1        4.4        2.9
            6.1        4.4        4.4
            5.9        4.6        5.9
            5.8        4.7        7.5
No          6.1        -11        2.9
PSV         6.1        -11        4.4
            5.9        -10        5.9
            5.8        -10        7.5
COPYRIGHT 2017 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc. (ASHRAE)
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Improving Thermal Loop Performance
Author:Chen, Hui; Riley, James; Williams, Les; Chen, Amy; Bruner, Homer, Jr.; Henry, Robert
Publication:ASHRAE Journal
Article Type:Report
Geographic Code:1U7TX
Date:Apr 1, 2017
Previous Article:Smart Design at Chillventa.
Next Article:Living While Learning.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters