As part of an energy reduction effort, a Canadian technical college hired a compressed air auditor to do a leakage audit of their large campus, which houses over 30 mixed use buildings, including laboratories, research facilities, shops and classrooms. The audit found very few leaks, the reduction of which would achieve minimal savings; however, a few surprising items of interest were noticed during the study that showed very good potential for operating cost savings of 64% with an estimated \$45,000 per year in reduced energy and water costs. This article discusses some of the findings and how savings can be achieved on lightly loaded compressed air systems.
Supplying Compressed Air to Multiple Buildings
The campus under study has been in existence since the 1960s and is now among the largest colleges of its type in Canada. More than 48,000 students are enrolled annually at the college. The courses offered are designed for the workplace, including degrees, diplomas and certificates spanning Applied and Natural Sciences, Business and Media, Computing and Information Technology, Engineering, Health Sciences and Trades, with most subjects requiring hands-on learning with compressed air-powered tools and machines. Also, on campus are applied research activities to help bring new products to the marketplace and address industry-specific problems; this function also requires compressed air to run assembly and testing apparatus.
The campus, and the original location of the school, started out small and expanded greatly over the years. One by one, as various areas of instruction were added, specific buildings designed for the instructional activities were constructed, the number now totaling over 30 separate buildings. In most of the buildings, among other utilities, there are at least one 100 psi compressed air system. The systems are made up of reciprocating, scroll, and lubricated screw air compressors, which were chosen to match the desired duty of the compressed air uses within the buildings. Most of the systems use refrigerated air dryers, but one main powerhouse system supplies buildings containing laboratories and research facilities with instrument-quality air processed by a desiccant air dryer.
Unlike industrial sites, where the air compressors supply typical eight-hour shift oriented on cyclical loads, perhaps five days per week, 52 weeks per year, this college operates more or less on a classroom cycle, based on the school semesters. This means high peaks might be experienced due to demand in training shops when practical hands-on learning activities are going on, but most of the time, when students are in the classrooms, or during evening and weekend hours, there is very little compressed air demand. During these downtimes the air compressors in this facility are still active, providing pressure to the buildings, yet feeding only compressed air operated HVAC controls and the small flow caused by leakage in the various distribution pipes, connected hoses, valves, fittings, and energized machines.
This light loading presents a problem in terms of efficiency for some of the installed air compressors. One of the characteristics of lubricated screw air compressors running in load/unload mode is that they typically need to run for a period of time in the unloaded conditions before turning off after a load cycle. This run time is chosen to prevent excessive starts that may burn out the air compressor motors, usually adjusted so that the unit never starts more than six times per hour. But in the unloaded condition air-cooled lubricated screw type air compressors will consume between 25 and 40 percent of full load power while producing no air, a very inefficient way to produce compressed air. Depending on the installation characteristics, usually the air compressor storage receiver being the most critical component, a lightly loaded screw air compressor may even spend most of its lifetime in the unloaded condition, wasting significant energy and wearing out the air compressor prematurely.
Use of Hour Meters Reveals Inefficiencies
As a part of the leakage study, the compressed air auditor requested a list of all the air compressors and dryers on campus. From this list particular systems were chosen based on the air compressor size and type, with the screw units being of the most interest. Since there were quite a few buildings with compressed air systems, and scheduled audit time was short, the auditor needed some way to prioritize the expenditure of time. Once the key air compressors were identified, the auditor next focused on the operating hours of the air compressors; loaded and running hours were taken by site staff and these were used to determine which campus compressed air systems were consuming the most air, and the most power.
Figure 1: Calculation of average power from hour meter readings can be inaccurate if a lubricated screw air compressor has limited control storage receiver capacity. Simple calculations use a straight-line approximation line as seen with the green line. But the actual power depends on the amount of storage and the width of the pressure band setting. For an air compressor with one gallon of storage per cfm running at 10% capacity the calculated power would be 30% of full load when the actual power would be 58%.
Most modern air compressors have internal timers within the onboard controls that track the number of hours the unit has been loaded and the total number of hours the air compressor has been running. Using these timer readings, spaced between certain known times and dates, a rough calculation of the air compressor flow output, energy consumption, average power and specific power can be calculated. This is like a mini-audit, only one using no measurement instruments.
An example of this method is as follows for the air compressor designated as GA18+FF on the campus. This unit is a 25-horsepower (18 kW) lubricated screw air compressor with an internal air dryer within the air compressor enclosure. This unit was found to have 1,672 loaded hours and 11,083 running hours in its lifetime when it was surveyed initially. This is the first hint of inefficiency. A screw air compressor with a loaded-to-running ratio of less than 10% will be very inefficient, however, an initial hour meter survey cannot be relied upon because these are lifetime hours, and the machine could have been running under different conditions, even in a different location through most of its previous life.
Taking the hour meters again exactly a week later confirmed the air compressor operating was inefficient. The new readings were 1,685 loaded hours and 11,218 running hours. Doing simple subtraction, we can calculate that during this period the air compressor was loaded for only 13 hours, yet it ran for 135 hours out of the 168 hours of elapsed time during the week. Subtracting loaded from running hours we can see that 122 hours of operation was spent in the unloaded condition where the air compressor was consuming power but producing no air.
From these hours we can do a rough estimate of the power consumption of the air compressor and the flow output. Typically, the full load power of an air compressor can be estimated by taking the nominal hp and multiplying by a factor of 0.85, in this case for a 25-hp unit the full load power would be estimated at 21 kW. For the unloaded condition, an air-cooled air compressor with fans would consume about 35% of full load power, or in this case about 7.4 kW. Screw air compressor flow output can be roughly estimated by multiplying the nominal hp by a factor of four, or in this case about 100 cfm (4 x 25 hp). For better estimates, or if an actual measurement is desired, consult the manufacturer. The calculations for this air compressor are as follows:
- Average energy at full load = FL kW x FL hrs = 21 kW x 13 hours = 273 kWh.
- Average energy at unload = 7.4 kW x 122 hrs = 903 kWh or 77% of the total energy in the period.
- Total energy = 273 + 903 = 1,176 kWh or an average of 7 kW over the 168-hour duration.
- Flow output = FL hrs / elapsed hrs x rated flow = 13/168 x 100 = 7.7 cfm.
- Specific power = kW/100 cfm = 7.0 / 7.7 x 100 = 90 kW/100 cfm.
Note these are only rough estimates and do not account for the fact that cycling losses could cause even more inefficiency if the air compressor was installed with a very small receiver. Figure 1 shows how the calculated value would be about 30% of full load at 10% loading, but the actual power consumption with one gallon per cfm storage size would be 58% of full load. This should be considered when interpreting this rough calculation.
The previous calculations show this GA18+FF air compressor is running very inefficiently (not even considering the internal non-cycling dryer power). A typical air compressor running optimally would have a specific power number, which is a rating somewhat like an air compressor “gas mileage rating,” of between about 20 and 25 kW per 100 cfm, depending on its discharge pressure and the size of the air compressor. Usually smaller air compressors have higher specific power levels that larger units, the rated specific power of most popular brands of air compressors can be found in the Compressed Air and Gas (CAGI) data sheet found on most manufacturer`s websites.
Specific Power Varies Widely
Figure 2 shows the survey results of some selected air compressors from the facilities. It can be seen that specific power varies widely for the various systems from a low of 20 for systems with VSD air compressors to a high of 90 kW/100 cfm for a lightly loaded unit with very small storage. Projecting similar operation over a period of a year (52 weeks) the estimated annual costs have been calculated in Figure 2. Some of the systems are operating at less \$350 in power costs per year, yet others are costing thousands of dollars per year. It can quickly be seen that spending leak detection hours on a system with an average flow of under four cfm and a good specific power would not be very fruitful. But addressing the air compressor control issues has a very big potential for savings.
Figure 2: Based on simple calculations using the loaded and running hours taken one week apart, the specific power (yellow), energy (kWh) and operating cost (orange) can be roughly calculated. These calculations allowed the auditor to focus on the systems with the highest cost.
Armed with this new knowledge the compressed air auditor focused his attention on the systems with the largest flows and annual costs. By far the largest one was in the campus central power house where an apparent flow of 158 cfm was being consumed by the system.
Examining Central Powerhouse System Opportunities
The campus has a central powerhouse where hot and chilled water are produced for use in the various buildings for heating, cooling and hot water. In the basement of the powerhouse are the main air compressors that feed the instrument air used by the powerhouse systems, and for use in laboratories in three connected buildings.
Two air compressors run in alternate duty, one 50-hp, air-cooled VSD unit runs in winter months, and a larger 75-hp, water-cooled screw air compressor running in load/unload mode operates in hotter summer months. Both feed into a 330-cfm rated heatless desiccant dryer with dewpoint dependent switching control. The dewpoint control is designed to reduce the constant 50 cfm purge of an uncontrolled dryer to a less costly level proportionate to the lower average loading of the dryer.
The VSD air compressor was purchased to be the lead unit, with the 75-hp unit intended to run as backup. Unfortunately, the basement location of the air compressors has very little ventilation causing the air-cooled VSD unit to overheat in warmer weather and trip off, so the large water-cooled air compressor is required to run. This larger machine is much less efficient and uses metered city water, which costs almost as much as the air compressor energy consumption.
The air dryer dewpoint dependent switching was found to be faulty. Even though the dryer was processing only 47% of its rated flow, the dewpoint on the dryer never reached the minus 40 oF rating of the dryer, the reading on the installed dewpoint meter showed between minus 9 oF and minus 15 oF. This high dewpoint meant the dryer continued to consume the 50 cfm rated purge flow, making the percentage purge contribution to the total compressed air flow 32%, rather than the normal 15 to 20 percent of a properly controlled desiccant dryer. The dryer desiccant had recently been changed to try to correct the problem. A dewpoint probe calibration problem is suspected.
Analysis of the rough calculations in Figure 2 shows continuous use of the VSD air compressor would be much desired since it has much better specific power and does not use expensive cooling water. In terms of leakage reduction, the remaining approximately 100 cfm of compressed air flows into the three connected laboratory buildings, representing the largest compressed air cost of the campus. A leakage survey of these buildings, however, found minimal leakage. It is suspected that the compressed airflow is consumed by hidden leakage, perhaps corroded piping, or through abandoned uses, or compressed air consumption on equipment not being used, but left pressurized.
Because of this significant but unknown use the auditor recommended further investigation, and the installation of flow meters for each building feed so the source of compressed air demand can be isolated. In this case the easiest compressed air waste to fix is the excessive air dryer purge.
The powerhouse is currently installing a heat recovery system on the VSD air compressor as an energy efficiency project. The air compressor manufacturer offers an oil-to-water heat exchanger that can transfer the heat of compression to a water flow. Since the powerhouse produces all the hot water for the campus, preheating the water is thought to be the perfect use for the heat, which will supplement the natural gas fuel. A side benefit of this heat recovery system is that it will take away most of the heat that is currently overheating the poorly ventilated air compressor room.
High Energy Costs with Internal Refrigerated Dryers
Most of the screw air compressors on site have internal refrigerated air dryers (dryers installed within the air compressor enclosure) running in non-cycling mode. Normally refrigerated air dryers consume only a small percentage of the total compressed air system energy, but for lightly loaded systems the total energy consumed by the dryer can even exceed the air compressor energy. Dryers installed inside the air compressor enclosures are typically slightly larger in capacity that normal external dryers, in order to deal with internal enclosure heat.
Figure 3 shows the estimated dryer cost compared to the cost of air compressor operation. This estimate confirms that four of the system consume more energy than the associated air compressor. Of the \$22,000 in annual operating cost for the smaller compressed air systems, about 22% of the energy is being consumed by non-cycling refrigerated dryers.
Figure 3: For air compressors with internal refrigerated dryers, the operating costs of the dryer (brown) sometimes exceeded the cost of operating the air compressors (orange). A significant operating cost with the water-cooled air compressor was the metered city water, which approached the total electrical cost of operating the air compressor.
The air compressor manufacturer does offer a cycling option for the internal dryers that could reduce the dryer cost if modifications can be made.
A leakage survey identified additional energy savings opportunities. In survey conducted using an ultrasonic leak detector in a select number of buildings found about 25 leaks. Based on 10 cents per kWh, the savings that could be gained if these were repaired (and the air compressor control optimized), resulting in a savings of approximately \$5,000 per year. This would be enough to pay for the cost of the leak assessment. This potential savings is small in comparison to the savings that could be achieved by improving the air compressor and air dryer control for the various systems.
Figure 4: A small number of leaks were found, accounting for about \$5,000 per year in electrical costs. By far the largest cause of leaks was poor sealant and loose fittings.
Recommendations for Approximately \$45,000 in Savings
In looking at the analysis in Figure 1 we find some of the air compressors have excellent specific power numbers, leading to reduced operating costs. This is a clue to the potential savings if the systems are to be improved.
Visual inspection of the systems, for example in two automotive shop buildings, showed large storage was installed that allowed the screw air compressors to run almost fully in start/stop mode, greatly reducing the specific power levels, and the operating costs. Adding storage to the poor performers is a simple solution to improve the efficiency of most of the screw air compressor systems. Some recommended solutions:
- Install large storage to enable start/stop operation.
- Convert air dryers to cycling mode where possible.
- Replace older air compressors with start/stop units with cycling dryers. It should be noted the reciprocating and scroll air compressors at this campus have no unloaded run time and would be a good choice for future replacements.
- Turn off the air compressor when classes are not in session.
- Correct the dewpoint-dependent switching on the desiccant dryer.
- Install heat recovery on powerhouse VSD or improve the ventilation.
- Continue leakage investigation on powerhouse system.
- Install an air compressor flow and energy monitoring to ensure system efficiency.
It is estimated that system improvements at this site could yield an operating cost saving of 64% estimated at \$45,000 per year in reduced energy and water costs.
This article shows how unexpected potential savings can be gained by allowing an experienced compressed air auditor to assess compressed air systems. Unexpected potential air compressor control and air dryer savings can often detect, and the possible savings for these far outweigh the small savings potential of a simple leakage survey. Often compressed air system operators are unaware their air compressors are wasting power in the unloaded condition, but there are ways to improve the situation.
For more information about this article, contact Ron Marshall, Marshall Compressed Air Consulting, tel: 204-806-2085, email: ronm@mts.net.
To read similar water savings articles visit www.airbestpractices.com/sustainability-projects/water-conservation.
