7. Whole-Building Integration

The line between system-level integration and integrating those systems with each other, the building, and the environment is not distinct. There are components in many of the examples and topics discussed in the previous section that touch on building and environmental issues in addition to the multi-system integration issues. The bottom line is that following the testing hierarchy depicted in Figure 4 will lead the commissioning process from a level where virtually no integration exists through layers of increasingly complex integration. The following section examines some of the issues and concepts associated with testing at a multi-system and whole building level.

1   System start-ups and shut downs are major events.

One of the most demanding things buildings and systems ever do is come up to speed in an orderly manner from a dead start or shut down. This often is at odds with one of the simplest energy conservation strategies out there; turning things off when they don’t need to be on.

When a system is started up or restarted (for example recovery from a power failure), a major step change is introduced into virtually every control loop in the system. Given the interactive nature of these processes, escalating instability and even damage to equipment can occur if adequate attention is not paid to such operational changes. The first step in preparing for start-up and restart events occurs during the design process by properly sizing and selecting equipment. Equally important is having a detailed sequence of operation that describes exactly how each system should be started or restarted. Refer to the Recovery from Power Failure Test Guidance to assist in understanding the steps necessary to bring an air handling unit back on-line after a power failure.

The first time a system experiences scheduled operation, the event should be approached with caution, especially in large, complex buildings and systems. This is because the events that will transpire may push the limits of the testing effort to date. Generally, it will be better to test one end user system, such as an air handling unit, and the response of its utility systems (such as chilled water, hot water, and steam) to scheduled operation before allowing all of the systems in a building to operate on a schedule. In fact, staggering the start-up of systems is a fundamental way to minimize the potential for problems.

It also is important to remember that problems will more likely occur during extreme conditions. Thus, a system that has been tested in the early or late summer and found to be satisfactory may experience problems during the first scheduled start-up in the fall or winter.

Putting a system through a scheduled start-up test is an excellent integration test, especially if weather conditions are extreme. It is not uncommon for commissioning providers to get a lot of pushback when they suggest performing a scheduled start-up, especially in large, complex facilities, and especially if the systems are specified to run “around the clock.” In fact, the magnitude of the pushback is a good gauge of the advisability of proceeding. If the operating or construction team is reluctant to allow the test, they probably know something you do not, and it would be good to understand the rationale for their concerns.

2   Systems that run continuously must eventually handle unscheduled shutdowns and restarts due to power failures, fire alarms, and equipment malfunctions.

Many systems are intended to operate 24 hours per day in order to serve various loads like computer and server rooms, process lines, or medical facilities just to name a few. The argument that there is no need to see how these systems respond to a shut down and restart because they are “specified to run around the clock” is shortsighted. The reason being that Mother Nature and Murphy’s law have little regard for project specifications, scopes of work, and contract boundaries. Generally, and without the consent of the owner or construction manager, they conduct an independent test of the building’s automatic shutdown and restart capability during the first year of operation via a power outage, false fire alarm, equipment failure, or similar event. If the systems cannot handle this contingency, the results can be disastrous. However, the fact that there is resistance to testing these strategies is a pretty good indicator that there are issues that need to be addressed. The Recovery from Power Failure Test guidance outlines the things that should be considered when developing such a test.

3   Central plant and system interactions.

Almost all end use systems, such as air handling systems, will interact with central plants like chilled water, hot water, and/or steam systems. And conversely, because the central plants are common to multiple end use systems, they often serve as a medium for propagating interactions between systems, to the detriment of all. A common example of such problems is illustrated in Figure 14. The AHU in this illustration is served by a variable flow, primary secondary chilled water plant. The unachievable discharge temperature set point causes the chilled water valve on the unit to drive 100% open and stay there, regardless of the actual load. As a result, the flow through the coil is constant, which negates the fundamental design principle of variable flow system – the flow will vary with the load.

Figure 14: An AHU set point problem that will impact the chilled water system which will in turn, impact the other AHUs on the site as well as the triggering AHU.

Click figure to display it as a PDF.

The excess flow created by the inappropriate control valve position will bypass the plant as illustrated in Figure 15. The illustration on the top shows how the plant is intended to run at 25% load. Flow only occurs through the active load and the extra flow moved by the evaporator pump recirculates in the bypass line to the return line. This lowers water temperature entering the chillers and causes them to unload. The illustration on the bottom shows what happens to the chilled-water plant in a situation similar to that depicted in Figure 16. In this case, the wide open coil bypasses chilled water through the inactive load. This cold water lowers the return temperature and the chiller runs unloaded, just as was the case when the bypass had reverse flow. The difference, however, is that the wide open valve on the inactive load creates a flow rate in the distribution system that is analogous to 75% load. The flow in the distribution loop exceeds that provided by the active chiller and this excess water flows through the bypass line the chiller plant and dilutes the chilled water supplied by the operating chiller, raising the supply temperature. The elevated supply temperature only makes things worse since the active load must now open its valve further to achieve the same result and starts a vicious circle commonly referred to as plant over flow. Operating the second chiller solves the high supply water temperature problem, but now two machines and their associated pumps and tower fans are running with the compressors loaded at 25% when one machine could have run loaded at 50% if things were working properly. One possible outcome is a loss of supply temperature control at the plant. The elevated supply temperatures leaving the plant will no longer satisfy the loads that are active, causing their valves to drive open to deliver more flow and the situation spirals out of control. These conditions can be partially mitigated by operating the second chiller, which prevents the problem from rippling out to other loads, but reduces the chiller plant efficiency because the added equipment will consume a lot of energy. O&M costs also go up over the long-term as the extra chiller and its auxiliary equipment accumulate unnecessary operating hours. Either way, the distribution pumps waste energy because they are moving more water than is required by the current condition.

Commissioning the system to interact as designed optimizes its performance. The control valves interact with the distribution pumps in accordance with the loads to pump only the required amount of water. The variable flow distribution loop interacts with the constant flow evaporator loop, so the chillers load and unload in response to the requirements, and the number of chillers required is never more than absolutely necessary for the current situation.

The bottom line is that the interactivity of HVAC systems, buildings, and people is grounded in the laws of physics. Commissioning can make this fact work by ensuring that the laws of physics that were applied in the design process are manifested in the actual systems, efficiently delivering comfort and performance over a long equipment life. Commissioning failure unleashes physics to wreak havoc on systems, people, the environment, and operating budgets.

Figure 15: Variable-flow chilled-water-plant operation.

Click figure to display it as a PDF.

1   The integrity of the building envelope can have a major impact on the performance of HVAC systems and equipment.

The building envelope is a critical component of the building’s HVAC systems. A building envelope starts at the foundation and ends with the roof, applying the laws of physics to form a three-dimensional boundary between the built and natural environments. Envelopes provide shelter from temperature extremes, rain, wind, and other weather; protection from air contaminants and projectiles; barriers to groundwater and surface water, and protection from unlawful activities. Envelopes also provide aesthetic qualities to the built environment, through their architectural design, glazing, and finish materials.

Envelopes may be static systems relative to mechanical HVAC systems, but they are much more complex than they might seem, and their role in how well HVAC systems perform must not be underestimated. Assessing and maintaining envelope integrity is a key component of the successful operation of a building. Failing to do so can lead to energy waste, structural damage, damage to finishes, and indoor air quality problems.

The following example will illustrate the interaction of the envelope with HVAC system operation. A 16 story courthouse would experience significant cold spots, drafts, and even some frozen pipes on the lower floors during winter due to suspected leaks in the envelope. Facility staff had identified several major breaches in the envelope which were repaired but the building still experienced comfort and control problems. A simplified test based on ASTM-E-779-99 Standard Test Method for Determining Air Leakage Rate by Fan Pressurization was performed in which all return fans were turned off, all relief and return dampers were closed, each HVAC unit brought in 100% outdoor air , and the supply fan speed was gradually increased until the lobby became positively pressurized. The test reveled an estimated leakage rate of 100,000 cfm and the lobby was almost positively pressurized. Facility staff have already located and corrected some of the larger leaks and the control sequences have been modified to keep the return fans off and allow excess building pressure to be relieved through the envelope rather than relief dampers. This helps minimize infiltration into the lower floors and reduce the heating and comfort problems previously experienced. For more details on this particular example as well as general discussion on envelope failures and a field technique for assessing envelope integrity, refer to Commissioning and Envelope Leakage; Using HVAC Operating Strategies to Meet Design and Construction Challenges from the 2004 ACEEE summer study proceedings.[15]

2   Building occupancy patterns and population levels can have a major impact on the performance of the HVAC system and its energy requirements.

HVAC systems are designed to accommodate the needs of occupants, both current and anticipated. In most situations, this will mean the capacity of a system targeted for peak, worst-case conditions will be excessive for the normal, day-to-day operating state. In many instances, efficiency is preserved by designing the system to vary its performance in harmony with the requirements of the load served. VAV fan systems and variable flow pumping systems are good examples of this.

Many times, there are parameters associated with the design that do not lend themselves to automatic adjustment. Minimum flow rates on terminal units are one example, although advances in demand controlled ventilation technology are enabling a greater degree of automation.

Commissioning can play a key role in both of these areas by fine tuning and adapting the systems to the as-built, constantly evolving nature of the loads. This can have a major impact on efficiency on a number of fronts as illustrated in Figure 16.

Figure 16: The impact of matching the performance of the system to real-time requirements of the load.

Click figure to display it as a PDF.

In this example, the design load condition for the interior office cubicle was based on the following assumptions: 24-hour-per-day operation; electronic office equipment operating at full-load nameplate amps; and full office occupancy (based on the number of chairs in the office). The design minimum flow rate was based on ventilation requirements for two occupants within the office. The fact was that the systems were scheduled, the electronics used significantly less than the nameplate amperage, and only one person occupied the office for part of the day.

Actual operating conditions impacted zone cooling requirements significantly – the real peak was much less than the design peak. In fact, the flow rate needed to satisfy zone cooling requirements was below the design minimum flow position most of the time. Hence, the VAV box stayed at the design minimum flow position and operated more like a constant volume reheat terminal than a variable air volume unit. The minimum flow rate was reduced to a value appropriate for one. Consequently, the terminal was able to function as a VAV unit and all loads were satisfied.[16]

In this case, the impact of the excessive minimum flow setting rippled out to multiple systems:

·       Supply fan energy was affected because the fans moved more air than necessary to satisfy zone loads.

·       Load on the supply fan also impacts the amount of heat the fan/motor adds to the supply air stream.[17]

·       Reheat energy was affected because the unnecessary cooling represented by the high minimum flow rates had to be offset by heating the air stream to maintain comfort.

·       HVAC process energy was affected because the minimum outdoor airflow settings at the air handling unit were higher than required.

·       Central hot water plant pumping energy was affected because the excessive reheat load meant more water was pumped than necessary.

·       The central hot water plant pumping energy also was impacted by the preheat load represented by the excess minimum outdoor airflow in the winter months.

·       Central hot water plant gas consumption was directly tied to the reheat load. Reducing the reheat load reduced the gas consumption.

·       By design, reheat energy is simply a false load on the cooling plant. Eliminating unnecessary reheat eliminated unnecessary load on the chilled water plant.

·       Reducing fan heat had a direct impact on the load on the central cooling plant.

·       Reducing the minimum flow at the air handling unit had a direct impact on the load on the central chilled water plant.

·       Reducing the unnecessary load on the chilled water plant reduced its pumping energy for reasons similar to those listed for the hot water plant.

·       Reducing load on the central chilled water plant directly reduced the energy consumed by the chillers and cooling towers.

For the project illustrated in Figure 16, there were hundreds of terminal units with minimum flow rates that had not been tuned to meet actual operating conditions. In fact, the building was being ventilated for approximately 8,000 people when the actual population was less than 2,000. It is important to understand that for this particular owner, the excess capacity did not represent a waste of capital. The use of their office space was constantly changing and evolving and the quality, capacity, and flexibility of the systems they had purchased gave them the capability to adapt quickly and efficiently when required. Retrocommissioning identified a problem and provided a mechanism for correcting it by tuning the system to the existing conditions and training the operators so they could re-tune the system as loads changed. The bottom line is that making small tuning adjustments at the end-use equipment to meet changes in use/occupancy/load can save energy on multiple fronts, compounding the savings achieved at the device itself.

3   Seasonal variations impact system performance and need to be accommodated by seasonal testing and an on-going commissioning process.

Because they are tied to occupancy and construction schedules, the functional testing associated with most new construction commissioning processes will occur during only one season. Retrocommissioning processes often have more flexibility, but still may face deadlines driven by contract or utility program requirements.

Trend analysis can be a powerful tool for real-time functional testing after a building and its systems are up and running. This has been demonstrated in several examples in this chapter. In this scenario, Mother Nature writes the test procedure and the control system and data loggers capture the test results for analysis by the commissioning and operating team. Trend analysis techniques can be found in many of the resources listed throughout the Functional Test Guide, many of which can be downloaded directly from www.cacx.org or www.peci.org.

Trend analysis techniques also lend themselves to ongoing commissioning processes, where continuously running trends can be accessed by operators as needed to assess performance, fine tune systems, or diagnose problems

Training can be a valuable resource to comprehensively supplement commissioning processes. In new construction, training imparted by the commissioning provider and manufacturers familiarizes operating staff with equipment and systems and helps prepare them for operating them. Training also prepares operators for problems that might come up during the first year of operation. Similarly, training by the provider and vendors associated with a retrocommissioning project can add depth the knowledge-base of the operating staff and help the benefits persist. Finally, a facilities manager or commissioning consultant engaged in an on-going commissioning effort can provide periodic training sessions targeted at current operating issues or at renewing familiarity with the existing equipment. These sessions can reinforce the lessons learned and bring new team members up to speed.

HVAC systems can react in unpredictable manner with each other, the environment, the building, and its occupants. The following examples illustrate integration issues associated with the Student Center example depicted throughout the guide.

7.1. Integrating Warm-up with Hot Water Pumping

The final integration of the components and subassemblies associated with the AHU1 discharge temperature control cycle came together nicely during the early part of August as the Student Center moved toward final acceptance and occupancy. Component, subassembly, and system operation and performance was verified to the extent possible given the climate in August, and the commissioning team now worked with the control contractor to record on-going trends. These trends would keep a running log of the performance of the building and would be essential tools for fine tuning the building and completing the testing process. They also would serve as a valuable resource for the operating team as they moved beyond the warranty year.

The data showed its usefulness one pleasant day in October. The operating team gave the lead provider a call to discuss a cold building problem they had encountered that morning. They were perplexed because they knew the heating system and warm-up cycle worked because they had participated in its testing. And, the system had performed flawlessly the past several days while the temperatures had been hovering between 40°F and 50°F. However, after a low in the 50’s overnight, the hot water pump was not operating on this particular morning and they were bombarded with cold calls.

The problem was temporarily resolved by manually overriding the pump to warm up the building; however, trend data documented the root cause of the problem. The cool days and 50°F overnight low caused space temperatures to drift back to the night setback limit. But as morning approached, the outside air temperature rose steadily with the passage of a front and reached 65°F by 5:45 a.m. The sequence of operations stated that the preheat coil and pump would be locked out whenever outdoor air temperature exceeded 65°F, based on the assumption that preheat would no longer be required above this temperature. A true enough statement in general if the building was already up to temperature, but obviously not so true on the day in question. Even though all aspects of the warm-up cycle had activated as it should per the sequences, there was no heat because a separate control loop prevented the preheat coil from operating. The problem was permanently solved by adding a line to the control code that allowed the preheat coil to work if a warm-up cycle was initiated.

7.2. Integrating AHU1 with the existing central plant

One building-level integration issue of crucial concern to the facility operators related to how the dynamics of the new hot and chilled water loads associated with the renovation would impact existing central plant operation. In particular, the impact large load swings would have on the heating plant. The central heating plant consisted primarily of two coal-fired boilers, and some gas-fired equipment. An upcoming project would replace one of the original units with a high tech, fluidized bed boiler, but the chain grate stoker on the remaining original boiler would continue to operate for the foreseeable future.

Chain grate stokers do not respond to large load changes quickly or effectively. The operators could vary the rate of coal fed to the stoker, but once the coal ignited, it was going to burn. A sudden, major increase or decrease in load could really upset plant operations, including over pressurization of the system, “hammer” in the feed water and condensate return systems, and general imbalance in the water chemistry. Successful integration of the new HVAC equipment with the existing central plant requires careful review, and possible revision, of existing control sequences to accommodate the new dynamics of the overall system.