HVAC Strategy: Go with the Flow

Energy consumption and overall building efficiency have always been high priorities for building owners, but with rising energy costs and utility deregulation, these concerns have a renewed importance. These developments are creating a plethora of opportunities, as well as pitfalls, with respect to purchasing power.

By PETER D. ZAK, P.E., Principal, Zak Engineering Group, LLC., Mequon, Wis. October 1, 2000

Energy consumption and overall building efficiency have always been high priorities for building owners, but with rising energy costs and utility deregulation, these concerns have a renewed importance. These developments are creating a plethora of opportunities, as well as pitfalls, with respect to purchasing power. The challenge for mechanical engineers is to design heating, ventilation and air-conditioning (HVAC) systems that accurately respond to heating and cooling loads while maintaining the highest efficiencies possible.

Generally speaking, the largest consumer of power in a building is the HVAC system, with just about the only consistent factor being that heating and cooling loads are inconsistent. To address this, load calculations are traditionally used to evaluate the building at the two worst outdoor design conditions for heating and cooling, assuming that internal loads remain constant. Equipment is then selected to meet these conditions with the understanding that as the external loads fluctuate the equipment will modulate to respond to the building requirements.

This method of design, although somewhat conservative, is based on static conditions, which assume that outdoor conditions will be the driving force with respect to heating and cooling loads. In reality, a building is a dynamic entity where heating and cooling loads may occur simultaneously. Factors such as building mass, material capacitance and fluctuating internal heat gains-combined with varying occupancy levels and changing weather conditions-all affect net HVAC system capacity. The system must respond to the resulting load by adding or removing heat to maintain the space-condition setpoint at the time of occurrence. It is rare for all spaces to be at maximum load at the same time, and all buildings have a diversified loading profile-a fact that should be considered when evaluating equipment and studying operating performance.

While heating and cooling equipment maximum efficiencies have been enhanced over the years, in most cases these improvements occur only at full load. At part-load conditions, the efficiency may drop to levels that increase energy consumption.

As a result, the presence of single or multiple high-efficiency components does not always ensure maximum system performance. Unless all components are selected to operate harmoniously, overall system performance will suffer.

Various computer programs are available that evaluate the dynamics of building loads and accurately determine heating and cooling requirements not only at design conditions, but throughout the part-load operating range.

Hydronic systems: Pump less

Hydronic heating and cooling systems have traditionally proven to be the most suitable choices-with respect to efficiency and flexibility-to meet diversified building loads. High-efficiency boilers and chillers are capable of meeting building loads throughout the full operating spectrum. Combining this equipment with variable-speed pumping systems can maximize the overall HVAC system efficiency by allowing the plant to accurately respond to diversified building loads.

A very wise engineer once stated that there are three objectives when designing an energy-efficient system: pump less, pump less and pump less. There is however, an addendum: think system efficiency, not component efficiency and integrate control functions. Simply stated, by controlling water flow, pump horsepower can be reduced and primary equipment sequencing can be optimized, which together further reduces the overall energy consumption of the HVAC system.

There are several common variable-pumping configurations to be aware of:

Each system has merit with respect to hydronic system design and performance. The final system selected is usually chosen based on the application and the desired operating and control strategy.

The basic theory of variable-flow systems is to maintain a fixed temperature difference. This can be illustrated using the formula British thermal units per hour (Btu/h) equals gallons per minute (gpm) times 500 times the change in temperature (Dt), or Btu/h = gpm x 500 x Dt.

In constant-flow systems, gpm is constant and Dt will change to meet the load, or Btu/h. Pump energy consumption remains constant and primary heating and cooling equipment modulates in response to the Dt required to meet the load. Properly selected chillers and boilers remain efficient at part-load conditions, minimizing energy consumption. However, if the system water-temperature difference remains constant and flow is modulated, energy consumption is further reduced.

Hot-, chilled- and condenser-water systems are all viable candidates for variable-flow applications, if properly controlled. For most applications, the pumps are selected to operate in a parallel configuration. Pump speed is controlled by system differential pressure and equipment is staged on or off in response to system flow rates.

A sample system

Figure 1 is a schematic of a chilled-water system using a basic constant primary flow with variable secondary flow. In this configuration the differential-pressure transmitter (DPT) is located across the cooling coils, and the automatic valves sense the pressure drop as they modulate in response to air-handling unit (AHU) loads.

To ensure proper sensing, the control pressure drop must be measured across the coil and the automatic valve. In systems with multiple branch circuits such as a campus or an airport, the DPTs can be strategically placed across supply and return headers to maintain the required differential pressure. The signals are then collected and compared at the pump-speed control panel. The pump speed is modulated to provide the required pressure at the DPT farthest off setpoint.

In the diagram, Flow Meter 1 (FM-1) measures the secondary-system water flow and sends a signal to the chiller sequence controller, which starts and stops the chillers and the primary pumps in sequence to match system load. In this example, the primary pumps are constant speed and should be sized to meet the largest temperature drop possible across the chiller without sacrificing capacity or efficiency.

Flow Meter 2 is used to measure flow in the bypass and serves as an indicator.

The secondary-system DPT measures the system pressure drop and is used in conjunction with FM-1 to calculate the actual water horsepower used. Actual (wire) power consumption can be measured by the kilowatt (kW) transmitters located on the variable-frequency drives that control pump speed. The consumption can then be compared to waterside power requirements using a software program that determines the optimal point to operate either a single pump or both pumps in parallel.

In this system, primary and secondary water mix in the bypass piping, increasing or decreasing the supply-water temperature to the AHUs. The design intent is to operate the lead chiller at or near its maximum capacity before starting another machine. Before a lag chiller is started, flow in the secondary loop should actually exceed the chiller primary flow rate for a predetermined time period. The reverse holds true for taking a chiller off line. This prevents unnecessary start-ups and shutdowns that can create unwanted spikes in electrical demand. A good starting point is 20 percent of chiller flow rate for 15 minutes. The purpose is to have chillers in a multiple-unit configuration stay at or above 50 percent of the unit capacity. Most chillers have an optimal efficiency “sweet spot” from 50-percent to 90-percent loading, where the kW/ton consumption is actually less than the full load kW/ton.

The success of this control scheme relies on several key factors. First, pump selection is critical; the pumps should be identical and system curves for both single and parallel operation must be calculated to establish the most efficient operation. Second, accurate sensors and software are essential to properly control the pumping system.

A common mistake when selecting pumps to operate in a parallel configuration is to simply divide the flow in half for the design pressure drop. A pump is then selected for this design duty, but a system curve comparing the operation of both pumps in parallel is not evaluated.

The schematic in Figure 1 assumes that each chiller requires 600 gpm and pumps 1, 2, and 3 are selected to operate two in parallel with the third pump for standby; therefore, each secondary loop pump has a flow capacity of 900 gpm. If a 100-foot pressure drop is assumed, Figure 2 (page 24) is then representative of the system curve for two pumps operating in parallel. Point A represents the design operating point of both pumps operating together, point B represents the condition for each pump operating in parallel and point C represents the operating condition of a single pump with respect to total system flow.

At point C, the pump impeller curve extends beyond the system curve. This is critical because if the impeller curve ends to the left of the system curve, the results will be improper flow, cavitation and potential pump damage. Although a single-pump system curve may appear to meet design conditions, composite curves for both pumps operating in parallel must be plotted to ensure that the system curve does not fall below the pump impeller curve.

Accurate pump data is crucial to determining the sequencing of the variable speed pumps. Using the pump affinity laws, when flow changes, pressure drop changes as a square root function and pump horsepower changes as a cube root function. Note that point C indicates a flow rate greater than 50 percent for the single operating pump.

This information is used by the variable-speed-pumping software to determine the optimum operating condition. An effective method used to determine proper sequencing is to evaluate the ratio of theoretical (waterside) power consumption and actual (wire) power consumption, known as the “best efficiency control” strategy. Pump and motor efficiencies are used in conjunction with the system load profile to calculate the system power requirements.

This design and control strategy allows the heating and cooling plant to properly respond to diversified building loads. As control valves modulate to maintain required space setpoints, the system pressure changes. As the system pressure changes, pump speed and flow are increased or decreased to maintain the setpoint. Heating and cooling equipment is sequenced on or off in response to system flow requirements.

As previously stated, the success of a variable-speed pumping system is not just component-based. An analysis and understanding of the building load profile is a key element. The equipment selected should be based on design conditions and part-load efficiencies and, when combined with proper system control, should effectively close the gap between

the building requirements and the primary heating and cooling equipment.

Harley Davidson University

Harley Davidson University (HDU) is an exceptional example of chiller plant production matching building load profiles. HDU is located in an existing 120,000-square-foot, six-story facility on the Harley Davidson corporate grounds in Milwaukee, Wisconsin.

This former manufacturing facility was completely renovated-inside and out-by The Kubala Washatko Architects, Inc., Cedarburg, Wis., and now provides hands-on training for Harley Davidson service technicians and dealers.

Building usage is divided into several key areas: a conference center with a capacity of 300 people; the Red Brick Cafe, a cafeteria for the entire corporate complex; laboratories complete with operational motorcycles for training; a mockup of a dealership; an archives area for Harley Davidson memorabilia; and the racing/research department.

The maximum peak-cooling load for this building is 375 tons. This assumes all spaces are at full occupancy on a design day-which occasionally does occur during special training sessions. A typical day, however, finds several hundred people moving about the building from laboratories to the conference center, cafeteria or other areas of the building-truly a fluctuating load.

The chilled-water plant as schematically described in Figure 3 (page 26) is a constant primary-flow with variable secondary-flow system. The plant has an ultimate design future capacity of 750 tons using twin-compressor screw machines that have a full-load rating of 0.625 kW/ton. This total capacity is expected to service the addition of future facilities adjacent to HDU.

The plant utilizes waterside economizers via plate-and-frame heat exchangers and cooling-tower water. During low outdoor wet-bulb conditions, the heat exchangers and towers are used to produce chilled water. In addition to this, return chilled water is precooled by the waterside economizer during light load conditions with low outdoor wet-bulb temperatures. This allows the chiller to unload and operate within the part-loading “sweet spot,” keeping kW/ton consumption at its absolute minimum.

The table above is a condensed load profile for HDU taken on a typical summer design day. As noted in the data, the effect of outside weather conditions is minimal.

The chiller plant is started at 6:00 a.m. to precool the spaces and remove the heat built up from the unoccupied period. The outdoor dampers are shut until occupancy starts at 7:00 a.m.

The building occupancy begins to taper off at 3:00 p.m., leading to shutdown at 5:00 p.m. Once the lights, people and equipment loads occur the system begins to approach a flat load profile. This flat load profile is the result of the variable-speed pumping system tracking the loads as they shift throughout the building. As additional areas become occupied and others are vacated, the corresponding AHUs respond to the load by opening the control valves, only allowing chilled-water flow where it is needed.

The philosophies of variable speed

Variable-speed pumping provides a means of reducing horsepower, but to truly realize maximum efficiency, the entire system must be evaluated and controlled as if it were a single component. Always keep in mind these three key design philosophies:

Wise engineers will remember that finesse, not force, is the key to system optimization.

Table – Outside Secondary Flow

TimeAir TemperatureGPM
6:00 a.m. 70 790
7:00 a.m. 70 400
9:00 a.m. 74 425
Noon 84 425
3:00 p.m. 88 475
4:00 p.m. 95 525
5:00 p.m. 94 350