Introduction
The following information is based on the assumption that the design of a building's mechanical system is appropriate for the building's specific application and will provide the desired environment if installed as designed, if controlled as designed, and if properly maintained. The people who have the greatest interest in the final outcome of any project, large or small, are those who will depend on the new or renovated systems. Therefore, in addition to being involved in the design process, the institution should become familiar with the procedures for construction, start-up/commissioning, and ongoing operation that apply to all building projects. This is especially true when strict environmental control is a high-priority goal.
The author has gathered the recommendations and information that follow after working through construction, start-up, commissioning, and subsequent problems on many projects with critical goals for temperature and humidity control. The findings may appear to come from a cynical engineer, but it is prudent to expect shortcuts from contractors implementing a professional's design rather than to assume the contractors will perform their jobs exactly as specified.
The suggestions herein are presented as verification tasks the institution will want performed. It is certainly not the institution's responsibility to carry out these tasks, although some may find it easier to do certain things themselves instead of convincing others to perform the tasks for them. Ideally, these tasks would be the direct responsibility of the designer, but it is a rare design team that assumes the level of detail described here.
"Construction Phase" tasks are almost always within the designer's scope of work, and it is safe to assume at least minimal attention will be paid to them. "Start-up/Commissioning" is often neglected, though not necessarily ignored, by the design team. It is not to the design team's advantage to find operational flaws in their system. Given the potential for conflict of interest problems, testing the installed system may be a task for a special consultant, if such a luxury can be afforded. Finally, the tasks under "Normal Operation" are almost never part of the design team's scope of work. Preventive maintenance and monitoring programs are typically the individual institution's responsibility and, unfortunately, are often considered late in the process, e.g., after a few months of operation when the system begins to degrade for lack of attention.
Construction Phase
The key activity during the construction phase of a project is to insure that the specified equipment is provided and properly installed. If it is not, it is imperative to insure that the items of equipment substituted as "equal" are indeed equivalent in quality and performance to those specified in the design documents.
Typically, "special" equipment that is not usually installed in standard commercial buildings will be provided by the contractor as specified. Humidifiers, dehumidifiers, activated carbon filters, etc. fall into this category.
Contractors will often want to use commonplace system components, such as air handling units, cooling coils, heating coils, fans, pumps, diffusers, dampers, control systems, etc. that differ from those specified. The contractor's motivation is typically economic (i.e., the substituted items are less expensive than the specified items), but it can also be that the contractor has more experience with, and therefore is more comfortable with, the substituted components.
It is usually the designer's responsibility to review the suggested substitutes and determine whether or not they are actually equivalent to the equipment specified. Designers vary widely in the amount of attention they give this task, but in general most are conscientious. The institution should be vigilant when working with design firms that separate the design and construction support functions between different departments. If the person/team reviewing the submitted substitutions is not the same person/team who specified the equipment, there may be a lack of communication regarding what specified characteristics are most important.
Equipment Characteristics
Some key specified equipment characteristics to insist upon are listed below:
Capacity. Will the substituted item provide the pumping, air distribution, heating, cooling, humidifying, dehumidifying, or filtering capability required? Cooling coils are particularly tricky to evaluate because of the differences between total cooling, sensible cooling (cooling with no dehumidification), and latent cooling (cooling with dehumidification) capacities. To insure proper dehumidification, which is often a misunderstood concept, one must see to it that the latent cooling capacity must be equal to or greater than that specified.
Size. Will the substituted item physically fit into the space allotted for it, or will the substitution require rearranging other components of the system?
Noise Levels. A number of different models for rotating equipment (such as fans) can usually perform the specified function. Different models (i.e., sizes) will generate different noise levels. As a rule, the smaller a fan wheel, the faster it must rotate in order to supply the same amount of air; the faster it rotates, the noisier it is. There is clearly a trade-off between size (and initial cost) and noise levels.
Reliability, Serviceability, and Support. It will be beneficial, throughout the life of the system, to consider the intangible characteristics of reliability, serviceability, and on-going support during construction. Beware of "off-brand" equipment manufactured by firms that may not remain in business over the next 20 to 40 years. It is critical to confirm that there are service companies within a reasonable distance of your facility who are familiar with the equipment installed. Otherwise, the equipment will operate as specified only until the first problem.
Equipment Installation. Proper installation of equipment should also be verified by the designers. Depending on the designers' contract, the number of site visits can vary from a total of two or three visits during construction to once a week or more frequently. Between visits construction work continues and is often permanently hidden from view behind walls or above ceilings before the designer returns. The enlightened client, who may be near the job site on a daily basis, will make a point of frequently walking around the site, visually inspecting the installation, and informing the designer of any anomalies discovered. The client will undoubtedly be a nuisance to the designer who is not accustomed to having “help,” but the client must live with the system after the designer moves on to other projects.
Installation Features Requiring Verification
Thermal Insulation and Vapor Barrier Integrity in Walls. The importance of insulation and a complete vapor barrier is indisputable. They must be installed properly to perform their functions. Contractors may take shortcuts, because once the finished wall is up it is extremely difficult to verify the existence of insulation and almost impossible, non-destructively, to verify the existence of a vapor barrier.
Ductwork. Significant deviations from the designer-specified ductwork sizes and routes can affect the capability of a system to function as required. Longer duct runs, smaller ducts, and more turns can all increase the total static pressure of a fan system. As static pressure increases, the amount of air the fan can distribute through the system decreases. As the air flow decreases, the system's ability to heat, cool, humidify, and dehumidify also decreases. Walk-through inspections should compare actual installation with designed installation. Duct sizes are not usually substituted, but note any changes in how the ductwork gets from the fan to the room-supply outlets and return inlets.
Turning Vanes. The absence of turning vanes in elbows will also increase system static pressure. Installation of vanes in 90° elbows is easy to verify from the exterior of the duct before the elbows are hidden above ceilings. The turning vane welds or screws shown in Figure 1 will be visible on both sides of the elbow. In addition to increasing system pressure, the absence of turning vanes will increase “air noise” from the ductwork.
Dampers. Manual volume dampers should be installed in all locations specified in the design. These dampers are used by the air balancer to insure that the proper amounts of air are delivered to each space, thus insuring desired temperature and humidity control. Volume dampers are also easy to identify from the exterior of the ductwork, because they have adjustment handles 
that protrude from the sheet metal as shown in Figure 1.
Ductwork Lining. Some systems use ductwork lining to serve the dual purpose of thermal insulation (which prevents heat loss and condensation on cold ductwork) and noise suppression. In order to verify that the lining has been installed, one has to actually find an opening in the ductwork. Openings can be found during construction at the ends of incomplete ductwork and in holes cut for supply and return registers.
Ductwork Insulation. If the ductwork is not internally lined, it should be externally wrapped with insulation and covered with an airtight vapor barrier. Verification of the vapor barrier is critical, because if humid air comes into contact with a cold duct, condensation will soak the insulation; this degrades its performance and causes a nasty mess.
Equipment Installation. Proper installation of equipment should also be verified by the designers. Depending on the designers' contract, the number of site visits can vary from a total of two or three visits during construction to once a week or more frequently. Between visits construction work continues and is often permanently hidden from view behind walls or above ceilings before the designer returns. The enlightened client, who may be near the job site on a daily basis, will make a point of frequently walking around the site, visually inspecting the installation, and informing the designer of any anomalies discovered. The client will undoubtedly be a nuisance to the designer who is not accustomed to having “help,” but the client must live with the system after the designer moves on to other projects.
Pressure Gauges. Finally, if specified, differential pressure gauges across air filters are important items that are sometimes neglected. These gauges provide a quick way of determining when filters should be changed.
Access Doors. Doors or panels should be installed to provide access to equipment requiring future maintenance or service. Sheet metal doors will be located in ductwork for access to coils, dampers, fans, humidifiers, etc. Architectural access panels should be installed in walls, ceilings, or floors concealing valves, motors, and any other moving equipment. They should also be provided at ductwork access doors.
Piping. Faulty piping is usually not a problem for long, because a leaky pipe will be noticed. Pipe insulation, however, can sometimes be forgotten, especially on pipe fittings such as elbows and valves. Missing insulation on cold pipes can result in condensation, which will drip from the pipes just like a leak.
Coil Monitors. If they are specified, it is important to verify that thermostats and pressure gauges at heating and cooling coils are installed. These items are invaluable when it comes to future trouble-shooting.
Controls. Most control system components are "invisible" to anyone casually inspecting a construction site, but the temperature and humidity sensors for spaces should be self-evident. Sensor locations should be verified and the cleanliness of the sensors protected. Sanding and painting in the vicinity of unprotected sensors should be prohibited, but if this is impossible to enforce, sensors can be temporarily covered with plastic and masking tape when necessary.
Change Orders. A final note of caution for the construction phase of a project has to do with change orders. Changes to a project after the "final" design documents are printed and distributed are expected on every job. Unfortunately, the distribution of change order requirements is often limited and many times the subcontractors actually affected by a change do not receive or properly incorporate the change order into their plans. Change orders are critical to the correct construction of a project (the designer will not take the trouble to issue change orders if they are not), and their implementation should be verified.
Start-up/Commissioning
At the end of the installation stage of a project everyone involved is fatigued and anxious to have it completed. Often there is no more money to spend, the completion is behind schedule, people need to move into the new space, and the designers and contractors want to move on to their next projects. This is exactly the time when a new burst of energy is required of someone, preferably the designer or other consultant familiar with the intended operation of the new system, to insure that the systems function properly.
The job is not complete until the systems are operating on a consistent basis, and it is the contractor's responsibility to make it so. Not many contractors would deliberately install a system which does not work, but many do not take the time at the end of a project to test their handiwork. In the contractor's opinion, there is no reason to believe a system will not function as intended, because the contractor has watched it being constructed day by day. Many an installation has been plagued by complaints from occupants from the day they move in because system verification testing was not performed.
Without verification testing by a professional unrelated to the contractors, the institution may need to call the contractors back frequently to "fix" the new system. The institution that does not know the intricacies of the system will be at the mercy of the contractors, who are not about to find something wrong with their own installation. The designers, probably long gone from the scene, will be blamed for a faulty system that the contractors are "doing their best to make work." The finger pointing will stop only when the institution gives up trying or calls in the designer or another professional consultant, who should have been involved immediately upon completion of the installation.
Historically the most unreliable parts of a new mechanical installation are air and water balancing and the automatic control system. Therefore these elements require the most attention during the start-up/commissioning phase of any project.
Balancing
Balancers are typically either subcontractors or regular employees of the mechanical contractor, and it is their responsibility to insure that the air supplied by fans is distributed to individual spaces in the quantities and proportions specified in the design. They also insure that water supplied by pumps is distributed to individual pieces of equipment as required to allow for the proper performance of that equipment.
Balancers use special instruments to test and measure air and water flow, and they should be required to submit a report to the designers following completion of the balancing procedures. Sometimes designers review only the report and agree that the flows recorded meet the requirements of the specifications. Designers usually do not consider it their responsibility to verify the results of the report. For this reason balancers have a reputation for recording the airflows required regardless of actual field conditions. This unfairly accuses honest balancers, but one should proceed under the assumption that the balancing reports are not 100% accurate.
Verification testing requires instruments similar to those used by the balancers and therefore requires an investment by the professional performing the tests. An institution can hire an independent balancing contractor with no vested interest in the results to perform verification testing. The results are likely to be much more reliable than the contractor's results.
Spot checking a few air diffusers and registers for comparison to the balancing report should provide an institution with a feel for the accuracy of the entire report. If the random samples agree with the report it is probably not necessary to test each air outlet and inlet. On the other hand, if the random sample results vary widely from the report, it will probably be necessary to check each device and include the new results in a report for presentation to the original balancer. The original balancer will be required to return to the site (at no extra compensation) to rebalance all systems and submit a revised report. It is hoped that the balancer will realize that the institution is "serious" about the balancing report (many institutions are not) and perform the work correctly the second time. The third-party balancer should still be called in to spot check revised reports until the institution is satisfied that the systems are balanced as specified.
The same is true for water balancing, although balancers tend to be conservative in initial water balancing procedures by providing more water to individual pieces of equipment than specified. This is preferable to insufficient flow because the standard use of control valves will automatically modulate the flow of water to equipment as required to achieve the desired environmental conditions. Although a precise water balance is more desirable, the results of an overly conservative balance will not be as detrimental to system performance as a poor air balance.
Automatic Controls
The brain of any mechanical system is its automatic control system. Testing of the controls, therefore, is critical to insure system conformance to specifications. Unfortunately this is another often neglected task, partly because many people do not adequately understand controls. If a project's designer does not feel comfortable performing the controls verification task, a third-party specialist should be brought in to perform the testing.
Each control system is different, especially with today's computer-based direct digital control (DDC) systems, but the designer or third-party specialist need not be fluent in the detailed programming and user interface procedures of the particular system being tested. The controls contractor should be present for the verification testing in order to perform the system-specific tasks dictated by the tester, and this requirement should be specified in the design documents.
The start-up/commissioning of the controls system will include the following three steps:
Calibration
All sensors, but especially temperature and relative humidity sensors, should be calibrated to insure that they are reading actual conditions. This process can be time consuming but straightforward with a sling psychrometer in the hands of an experienced user.
Whenever air flows are being controlled against setpoint quantities for pressurization or indoor air quality purposes, it is imperative that the air-flow sensor be calibrated in its installed location. This will require the cooperation of the air balancer to provide the actual air-flow readings against sensor output.
Automatic dampers need to be calibrated to insure that their positions are those required by the control system. If outside air dampers are to be set at a specified minimum during occupied hours to insure proper ventilation, the air balancer will again need to be consulted to determine exactly what damper position corresponds to the desired outside air flow.
Automatic valves for heating, cooling, and humidification processes also need to be calibrated to insure that their positions are those required by the control system. It is also important to coordinate the operation of different valves to insure that simultaneous heating and cooling do not occur unless specifically required by the control system.
Testing
The testing procedure involves "exercising" the control system components. This includes changing setpoints and physically watching valves and dampers modulate. The tests should also override parameters such as time of day and occupancy mode so that testers can observe fans starting, stopping, or changing speed. A test procedure can be developed for each control system strategy specified in the design to insure its proper implementation under most conditions.
By putting a system through its paces during the commissioning process, an institution is less likely to be plagued by unexpected control-system behavior after occupying the building. This is especially true of control systems that are commissioned during the summer and have not been operated during the winter. Summer operation may be acceptable, but without verification testing, there is no way to tell what is going to occur when the weather changes. By that time the controls contractor may be long gone and has no financial motivation for performing additional work on the system. Obviously, the same problems apply to systems commissioned in the winter and never properly tested for summer operation.
An unexpected control system malfunction could be disastrous to collections. Therefore it is imperative to subject the system to all conditions, whether real or artificially simulated, prior to occupancy of the building.
Fine Tuning
The final commissioning task is the fine tuning of the control system. This involves adjusting control-system parameters as required to achieve the desired accuracy and speed of response. Again, because each control system is unique, the actual work should be performed by the controls contractor, but the designer should verify the results.
In simple terms, the control of a single device, such as a hot-water valve, boils down to a mathematical formula with a number of parameters that can be changed to achieve different performance characteristics. The control system receives an input signal from a temperature sensor, for example, and compares that signal against the desired setpoint value for that sensor. If the signal indicates that the actual temperature is lower than the setpoint, the control system sends an output signal to the hot water valve that forces the valve to open a certain amount to provide more heat.
The graphs in Figure 2 plot an input signal (temperature) on the vertical axis against time on the horizontal axis for different values of the control formula parameters. Control Curve #1 shows a fast response, with undesirable large fluctuations around the setpoint temperature. Control Curve #2 shows a formula that eliminates fluctuation around the setpoint, but produces an extremely slow response. Control Curve #3 shows an “ideal” control formula that provides a fast response with no need for corrections (overshoot). Overshoot describes the creation of a condition that is too extreme, to which the system responds with a correction that is too radical. In this situation conditions bounce back and forth between extremes before gradually reaching the desired value, as in Control Curve #1.
The design documents should specify the acceptable limits of overshoot and undershoot, eg., ±1°F, ±3% RH, etc. It is the control contractor's responsibility to determine the parameters required to achieve the fastest response within those limits. This is often a painstaking, time-consuming, trial-and-error process. Many contractors will fly through using default rule-of-thumb parameters. These may be acceptable for standard commercial buildings where tight environmental control is not critical, but contractors must understand that they will be required to optimize their controls for a museum or archives application.
Normal Operation
Once there is a newly tested fine-tuned mechanical system functioning as designed, the institution is on its own. The institution may feel particularly poor after the capital-improvements budget has been exhausted, but this is no time to skimp. In order to insure the continued proper operation of the new system and the longest life possible for its components, the institution must pay attention to the continuing performance of the system. The level of attention required will be dependent on the complexity of the system, but even the simplest arrangements of equipment and controls will need periodic preventive maintenance, cleaning, and calibration.
Left to its own devices, a system will appear to operate just fine until a catastrophic failure occurs or until the environment in one or more spaces falls so far away from the desired setpoints that an institution has no choice but to notice. By that time, the required cure is likely to involve a significant effort and expense, because there may be multiple problems in the system. Add to this the fact that the expense is totally unexpected and the money is not available to return the system to its original "as designed" operating state, and one faces the prospect of losing the benefit of the building or renovation project. In order to avoid this outcome, money needs to be budgeted for proper maintenance of the new equipment.
One of the most cost-effective methods of achieving continuous peak performance from a mechanical system is to have at least one knowledgeable and trustworthy person who will take responsibility for the equipment. This person, whom we shall call the Mechanical System Coordinator (MSC), can be on staff, a consultant, a service contractor, or a combination of the three. The job is the same: regular monitoring, maintaining, and servicing of the mechanical systems. Except in the largest installations, this should not require a full-time person.
A good maintenance program will anticipate problems before they reach the critical stage. This can be accomplished by monitoring, at a minimum, the following system characteristics:
Monitor Space Temperature and Relative Humidity
This task must be performed in any facility requiring tightly controlled environmental conditions. The MSC should review these records frequently, looking for trends indicative of degrading system performance. Armed with this information, the MSC can have filters changed, coils cleaned, controls recalibrated, etc. prior to the development of conditions that are unacceptable and perhaps harmful to the contents of a space.
Monitor Utility Bills
The MSC should be provided with all electrical, gas, and oil bills as they are received by the facility. By tracking energy consumption over months and years, the MSC will become familiar with what is "normal" and will quickly identify anomalies that may be evidence of an underlying equipment problem. Anomalies can be investigated and solved before the environment being controlled is affected.
Monitor Filter Status
By keeping a constant awareness of the condition of both particle filters and gaseous pollutant filters, the MSC will know exactly when each type of filter requires replacement. Particle filter monitoring is straightforward and simply requires the installation and regular monitoring of a differential pressure gauge across each bank of filters. As the filter collects dirt, it becomes more and more difficult for air to pass through, causing the pressure to drop across the filters.
Particle Filters
If no more definitive information is available, the differential pressure limit for particle filters can be the manufacturer's recommended maximum, but it is best to learn what pressure drop was assumed by the designer for each bank of filters. The manufacturer's recommended maximums are typically quite high, and the designers probably did not assume all the filters were that dirty when specifying fan size. When the pressure drop across filters rises above the designer's maximum level, the amount of air distributed to the system decreases and, therefore, inhibits the ability of the system to heat, cool, humidify, and dehumidify.
There is another reason it is undesirable to wait until filters are extremely dirty before changing them. Under those circumstances dirt may fall off the filters into the ductwork during filter changeout and be carried into the conditioned spaces. Clearly this should be avoided. Depending on the type of filter, location of the facility, and the ambient air quality, particle filters can be expected to be effective for a duration of three to six months.
Gaseous Pollutant Filters
Monitoring and maintenance of gaseous pollutant filters is more complex and will depend on the type of filter in use.
Standard carbon filter trays require that a sample of the carbon be tested periodically, usually by the manufacturer, to determine its life expectancy. When the carbon is "spent," i.e., has absorbed as much contaminant as possible, the carbon in the trays must be replaced. Other types of filters will require different procedures to determine when they should be replaced or replenished. None of them is as simple as the particle filter procedure. Gaseous pollutant filters typically last for a minimum of one year and often much longer, depending on the ambient air quality and the particular pollutants being absorbed.
Monitor Control System Operation
The automatic control system also requires regular attention to insure that it continues to operate as designed. It is essential that the MSC be familiar with control system basics, but he or she does not have to be an expert in the programming and adjustment of the controllers. The MSC needs to know only enough to identify and intelligently communicate perceived problems to the original controls contractor or other service contractor. Ongoing control system monitoring tasks are similar to those performed during the start-up/commissioning phase, i.e., checking the calibration of sensors and verifying proper operation of all devices. Problems are often identified when the conditioned environment degrades, but this may be too late for some collections. The emphasis needs to be on preventive, not reactive, maintenance.
Renovations
There is one final note for the "Normal Operation" phase of a system, and that applies to subsequent space and/or system modifications. It can be safely assumed that space use and configuration will change many times before a building or mechanical system is replaced or comprehensively upgraded. These alterations must be approached with caution to insure that the original performance of the mechanical system is not sacrificed.
The mechanical system will need to be modified to accommodate most architectural changes, but this fact is often neglected by the people planning a "small" renovation. Mechanical equipment is usually out of sight and, therefore, out of mind. This problem is most likely to be avoided when there is a Mechanical Systems Coordinator looking after the equipment. The MSC should be consulted to determine what affect the proposed changes will have on the system, how the system can serve the new spaces, and whether or not further expertise, e.g., design engineers, will be required for the renovation.
Conclusion
In summary, the people who have the greatest interest in the final outcome of any project are those who will use and depend on the new or renovated building and mechanical systems. It is advantageous for them to be involved in the design process, and to become familiar with the construction, start-up/commissioning, and normal operation processes inherent in all building projects. It may not be the institution's responsibility to perform all these tasks, but it is a good idea to be cognizant of how an individual project is progressing and to ask the right questions at the right time as "reminders" to those who do have responsibility.
The tasks discussed here apply to all building projects, not just to major new construction. They should be repeated for each subsequent renovation, no matter how small. In fact, it may be even more important that the institution be actively involved in "minor" renovations, because the smaller jobs are the ones most likely to be taken less seriously by designers and contractors.
While each building project is unique, with its own set of problems and constraints, this paper's suggestions are universal in their application. It is important to understand the construction process and what you have a right to expect as a future occupant of the space. Armed with this information, any institution can intelligently insist that systems be installed and operated as designed.
Sources of Equipment and Services
Sources of equipment will depend on the specifications and experience of the designer. It is important to confirm the continued availability of local service and support for equipment components. The length of time a company has been in business, their rating with the Better Business Bureau, and designer and contractor experience with the product should provide some guides. It is always a good idea to ask for references to other institutions where similar equipment or systems have been installed, and to follow up with a call to these institutions.
Another institution with experience in a similar project can be asked to recommend a third party for verification and testing. A regional museum association or an experienced conservator may also have good suggestions.
Written by Rebecca Ellis, PE
Attribution-NonCommercial-NoDerivs
CC BY-NC-ND