When it comes to creating central plants, design engineers are faced with a long list of complications. Energy and construction costs are rising. Higher ventilation rates, large computer loads and the desire for more natural light has imposed increasing demands on mechanical systems. Environmental issues regarding building components and refrigerants are of great concern as well.
When it comes to creating central plants, design engineers are faced with a long list of complications. Energy and construction costs are rising. Higher ventilation rates, large computer loads and the desire for more natural light has imposed increasing demands on mechanical systems. Environmental issues regarding building components and refrigerants are of great concern as well. Facilities are displaying a larger variance in occupancy schedules. Now more than ever, central plant efficiency has become crucial, not only from an operational standpoint, but for cost and energy conservation as well.
The measure of success for any power plant is overall system efficiency. A common misconception is to single out a specific piece of primary equipment and assume overall plant efficiency is a reflection of this single component. All too often, compromises are made on other vital components, such as controls and instrumentation, in order to stay within budget constraints. In a worst-case scenario, with minimized instrumentation for system monitoring and control, the building operator may never know there is a problem until a complete failure occurs. This "maintenance-by-failure" approach gives the impression of a properly operating system, but it is certainly a disservice to an owner who must cope with system gremlins and high operating costs.
While it may be true that primary equipment—such as a chiller—will be the largest energy consumer, if auxiliary equipment such as pumps and cooling towers are not properly matched or integrated with the control functions of the chiller and the building automation system (BAS), system inefficiencies are inevitable. But with a little effort and planning, the power plant can be designed to meet current requirements and accommodate future expansions while maximizing system efficiency.
A change of vernacular
The key issues to address at the outset are the owner's expectations, the established budgets and the sophistication level of the plant-operating personnel. With this information and a clear vision of system requirements, the design engineer can begin the process of determining the most cost-effective solution.
This process begins with load calculations. There are two types of loading conditions: peak loads , which occur when all building loads reach the maximum condition simultaneously, and block loads , when all loads do not occur simultaneously. The latter is more common due to the inherent diversity of building conditions.
The building-load profile is used as the base line to select the primary heating and cooling equipment. While the natural tendency is to specify equipment with the highest efficiency at full load, it is important to make note of part-load efficiencies and the anticipated proportion of time operating at part-load. Additionally, all equipment—from primary equipment to pumps, fans and cooling towers—should be evaluated together at both full and part load.
Following this evaluation, the creation of a system schematic that identifies components, fluid flows, valve positions and sensing/control points should define the design intent for both installers and the building-controls contractor. Using the schematic, control sequences can be prepared and control points identified. The instrumentation necessary to achieve proper sensing and control should also be gleaned from this information. This will allow the BAS software to properly control and monitor the system operation, and two things should become obvious: the plant operator will be able to fine-tune the system in response to shifting load characteristics, and major component failures can be avoided by the early detection of control settings on borderline valves. Once the control sequences and instrumentation are properly integrated and monitored in realtime, the phrase "maintenance by failure" can be removed from the day-to-day vernacular.
At the completion of the project, and prior to occupancy, the system should be balanced and go through a commissioning process that identifies and corrects any operational deficiencies.
An example of proper system planning, instrumentation, control and operation can be found at General Mitchell International Airport in Milwaukee. In 1999, the airport began an extensive expansion project that included the addition of a parking structure, as well as modifications to access and service roads.
It also meant the demolition of the existing power plant and the creation of a new one—a decision based on a study that identified numerous operational problems at the existing plant. The study also noted that the existing plant—which consisted of three 500-ton chillers and three 15,000-MBh hot-water boilers—could not be cost effectively modified to accommodate future concourse expansion projects.
The pre-existing configuration distributed water through a constant-flow primary pumping circuit, with secondary pumps arranged in a bridge configuration throughout the terminal areas. This system was fundamentally flawed, as even on the worst design day only 1,100 tons of cooling were required. These loads were difficult to satisfy due to the imbalanced flows created by the bridge circuits. As a result, the chilled-water system operated as a peak system—which caused unwanted blending and meant that the supply water delivered to the air-handling units was warmer than preferred. The hot-water system had a similar problem: all three boilers continuously operated at part load to meet the demands of the constant-flow pumping arrangement.
Additionally, without the proper sensing and controls in place, the staff had been manually staging pumps from smaller units to larger units on the primary circuits to offset the problems created by the water imbalance and by the three-way automatic valves at the air-handling units.
For now and the future
The new power plant—which went into operation in 2001—consists of four 500-ton centrifugal chillers and three 15,000-MBh hot-water boilers. The chilled-water, condenser-water and hot-water systems all utilize primary variable-flow pumping strategies, with the entire system monitored by a BAS.
To accommodate future expansions, the plant includes space for two additional 500-ton chillers and one additional 15,000-MBh boiler. All piping, valves, control sensors and sequences for the future primary equipment were also set in place during the initial construction. This will allow the airport to upgrade as needed with a minimal amount of disruption.
The method of control involves several components structured around the BAS.
The variable-speed system controls chiller/boiler staging, as well as chilled-, hot- and condenser-water flow rates.
The pumping-system software modulates pump speed in response to multiple differential pressure sensors located at the furthest points of the distribution piping.
Based on measured flow, chillers and boilers are sequenced to maintain the appropriate water temperature.
The chillers and boilers operate in response to factory-supplied control software, and control panels are wired to the BAS so the facilities engineer can monitor alarm status and equipment diagnostics.
While the BAS is providing start/stop signals, the actual changes in setpoints must occur at the equipment control panel. This method proved to be a cost-effective solution for several reasons. First, the plant is operated 24/7 and staffed with well-trained operations personnel. Second, using the capabilities of packaged controls reduced the first cost of the system. The final key is the control logic that is used to modulate pump speeds in response to heating or cooling demands.
The success of this system has reduced overall pumping requirements by approximately 500 horsepower, which, according to studies, should result in an annual operating-cost reduction of $140,000.
The secret to optimizing power plant operation is carefully matching up the equipment and understanding the limitations of the software. Once this criteria is established, the engineer can coordinate the required interfaces to allow all components to operate harmoniously.
The ultimate goal is to provide the facility operator with a tool that effectively monitors and controls the system. If this happens, the result will be an optimized, efficient plant with a reduction of system shut downs and maintenance.
Even with the best-planned design, confusion can occur at the communication level of control devices. For example, the packaged operating software for the chiller or boiler may not readily integrate with the building-automation system (BAS).
One method of resolving this problem is to have the BAS itself provide start/stop and alarm monitoring, with the equipment using the factory-installed control software. In this situation, the BAS is not truly integrated with the equipment controls; usually it is just hardwired to an interface within the equipment control panel. Actual system control must take place at the equipment panel, and the BAS can only initiate a start/stop function and can monitor or see all the readings taken by the equipment software. Actual control of the equipment by the BAS does not occur.
Another possible solution is to have the building-controls contractor provide a duplicate set of sensing and control points, which are hardwired back to the BAS. This method is often used where an open protocol is not available to integrate the system control functions. The BAS software controls the equipment subject to the limits of the equipment safety controls. One drawback is that this is expensive and, in some cases, the equipment manufacturer may not honor the warranty—depending on the level of control attempted.
The most reasonable solution, however, is to require that the equipment manufacturer and the building-controls contractor communicate through an open protocol system such as the BACnet interface. The objective of an open protocol is to provide a common communication language so that all control components can accomplish the control functions through the BAS software. And although the BACnet methodology has not been completely adapted by every equipment manufacturer, it is rapidly becoming an accepted industry standard.