Controlling Demand in Compressed Air Systems

by R. Scot Foss, Plant Air Technology

Half of the energy used by most systems is wasted by uncontrolled demand.  Attention to system control and balance can pay significant dividends – more than $1,390/per shift/year in a typical uncontrolled 25 hp system.

Over the past 15 to 20 years, responsibility for the end results in compressed air systems has been shifting to compressor room equipment.  Driven by the development of packaged air compressors, this approach is responsible for considerable waste of energy and a number of consequential problems.

The desired result of a compressed air system is pneumatic power at the point of use delivered at sufficient volume and pressure to do the job.  When difficu9lties occur most owner-operators throw prepackaged engineered solutions at poorly defined problems with little thought given to configuration technology.  The pre-packaged solutions seldom produce more than marginal improvement for short periods of time.

Most perceived system problems can be solved by overwhelming them with energy.  However, this solution is certainly the most costly, from both capital and operating viewpoints.  Interestingly, it is not the user who believes he has problems that is pleased by such a “solution”, but the one who has “solved” these problems by wasting 30 to 50% of his input energy.  He feels pleased because management has supported him with capital to overpower the problems  This approach takes away one problem and leaves others.

The common requirement of 100 cfm at 80 psig is an expression of what is desired at this point of use. If all of the responsibility for that result is assigned to compress0or room equipment you have joined the multitude of users who have begun the endless journey of filling the bottomless pit of uncontrolled demand.

All compressed air systems have three fundamental elements:

·         Demand,

·         Distribution, and

·         Supply.

These three factors must be controlled for the system to work at optimum energy and quality levels.  A properly designed and operated system can be described as “refining the energy response to a controlled demand through a controlled storage system.

Only half of the demand in a typical uncontrolled system is the result of real production demand.  The rest goes for artificial demand, poor applications, and leaks, as illustrated in the accompanying chart.

·       Artificial demand, which represents at least 15% of typical system demand, is generated by an application where the operator has adjusted the pressure to a higher level than necessary (often wide open) in lieu of appropriate control pressure maintenance.  It is also representative of applications where a regulator was not installed because it was not considered essential to the application.  Regulators, however, are essential to system control because demand is a function of supply pressure.  A demand of 100 cfm at 80 psig, if not regulated at 80 psig, will increase to 120 cfm at 100 psig.  If a 100 cfm compressor at 100 psig is installed to handle 100 cfm at 80 psig, it cannot deliver.  The pressure will drop to less than 80 psi, and the system will use 33% more energy than necessary to do a poor job.

·     Demand from poor applications is generated by using compressed air for keeping workers cool during the summer, open lines for parts blow off, cleaning the floor and other inappropriate applications.  Most of these applications require impingement energy, calculated as:  Impingement energy = ½ mass x velocity².  Mass is a function of volume, so impingement energy is a function of half the volume.  However, impingement energy is a function of the square of the velocity.  Therefore, high-velocity or high-thrust nozzles can reduce the demand for these applications by 60 or 70% and improve the end result.  These devices, like all air equipment, must be applied rather than “thrown” at the problem.

·   Leaks are an ever-present problem for all users.  They are never dealt with because they are not identified by location, quantified by volume and pressure, or expressed in dollar cost.  When they are properly identified, management will respond quickly.  There are three types of leaks:

§  Abandoned equipment leaks where operators walk away from their work stations,    leaving the compressed air on,

§    Mechanical operational leaks in valves and controls requiring maintenance, and

§    Plumbing leaks in pipe, hose, disconnects, and fittings.

·    Both ends of the distribution system must be controlled.  At the points of use, demand is controlled by automatic filter-regulators.  At the other end of the distribution system, the discharge from the compressor room should be controlled with an intermediate mass flow controller or sector control.  All capacity to store air between these two locations is a function of the controlled differential pressure and the compressor’s operating pressure.

For every barometric pressure ratio (14.5 psia, for example), one capacity of air can be stored.  For example, a 400 gal. Tank has 53.5 ft³ of physical storage capacity.  If the difference between the maximum control pressure at the point of use (80 psig) and the intermediate control pressure upstream of the tank (90 psig) is 10 psid, the controlled differential pressure is 10 pdid / 14.5 psia barometric pressure = 0.69 of physical capacity.  This ratio represents 0.69 x 53.5 = 38.9 ft³ of storage at 90 psig.  In mass units, the storage capacity is 38.9 psig = 19.7 lb. of air in storage.

Typically, there is a little storage value in piping for meeting demand peaks, but each system should be evaluated on its own merit.  Dedicated storage should be used for high surge applications.  Check valves upstream of the dedicated storage isolate the surge load from the rest of the system.  Several intermediate controls should be used as sector controllers if there are sectors in the system such as various buildings, different pressure requirements, or areas where usage must be determined for accounting. Purposes.

Point of use controls and intermediate controls are essential to controlling storage, which in turn increases the capacity of the system and reduces the brake horsepower needed to power it.  Storage also affects the load/unload cycle of the compressors.  The more storage, the longer the load mode and the longer the unload mode.  Reduced cycling will significantly extend the useful life of the equipment even if the output is the same.

Without storage, compressors must serve every peak and valley of demand.  If compressors are set to supply the highest peak, the machines and accessory equipment will run during non-peak demand and thereby waste energy.  Storage could be used to handle the peak and keep the compressor off.  With pressure only controls on the compressor, it is not uncommon to have 30 cfm of added demand drop the pressure below the set point for two 200 hp compressors and load a third.

As can be seen in the second example system diagram, intermediate controls are the first priority for balancing the system, managing demand (including leaks), controlling storage, and unloading horsepower.  Control of the intermediate pressure and the compressor control pressure can change the storage and the weight flow of demand and significantly change the way the system operates.

Intermediate control is the most important control point in the system.  Because the system can tolerate no hysteresis or failure, the controller should be a proven high-reliability unit that can analyze upstream and downstream conditions and control both. This control point is an excellent place to evaluate flow, dew point, storage volume, and minimum and maximum pressures.  The controller should have tamperproof controls and failsafe circuits.  Without this feature, a control failure would cause a system failure.

The third system example shows the effects of intermediate controls plus improved point-of-use demand and leak repair.  Intermediate controls will affect all downstream demand regardless of the presence of demand controls and possible operator tampering to increase the maximum control pressure.  The controls must create a pressure differential.  The intermediate pressure must be lower than the lowest compressor pressure for the system to work properly.

The fourth system example of the series illustrates a fully balanced system, with automatic demand controls installed at all points of use to control the maximum pressure at which air can be removed.  This point-of-use control will create storage in the header, sub-header, and branch piping.

The example shows that only 5.25 ft³ of storage will reduce horsepower.  Therefore, the reduction in overall storage is not particularly critical.  The 2 psid pressure drop between the intermediate control and the demand should be eliminated.  And enough storage should be created in the piping to prevent draw down from surge at the point of use.  To accomplish this, intermediate control pressure must be increased, which will reduce the tank storage by 31 ft³.

This condition will not improve or reduce horsepower at the compressor, but it will shorten the loading cycles.  To lengthen the cycles, another storage vessel could be added and the compressor pressure maintained at the current level, or the compressor pressure could be increased (if possible).  The final balance of the example system does not seem to be a problem.  Obviously, a considerably smaller compressor could be used if it could run flat out.

The advantage of control is demonstrated by the last example, which compares two systems featuring identical equipment, piping, and demand applications, except that one system has intermediate and demand control.  System A is a typical uncontrolled system.  System B is properly configured with control.  Point-of-use regulators alone do n ot necessarily indicate that demand is being controlled at that point.  All points of use must be controlled to a maximum pressure lower than the lowest compressor pressure to have balance.  As the demand pressure rises, so also does the demand.

Without controls, demand in System A increased to 131 cfm at 103 psig, exceeding the capacity of the compressor.  As a result the point-of-use pressure will fall as demand exceeds the capacity of the compressor.  The compressor will run flat out and create 131 cfm at 78.9 psig at the point of use as a function of the initial supply pressure to a demand, which if controlled at 80 psig would have consumed 100 cfm.

The point of use and intermediate controls in System B will maintain demand at 100 cfm or 48.24 lb.  Assuming that at least 25% of the demand has a use factor of 50% or less, 12.5 ft³ of the useful storage will reduce what the compressor “sees” as demand (12.5 ft³ at 80 psig is 6.03 lb.).  The compressor will interpret demand as 48.24 lb – 6.03 lb = 42.21 lb. or 87.5 cfm at 80 psig, or 66.47 cfm at 110 psig.  The weight flow of demand has now been balanced with the weight flow of supply interpreted through storage.

The next problem is to reduce electric power to match requirements.  System B has a perceived demand equivalent that is about two-thirds of the full load power of the compressor.  An equivalent amount of kilowatts should be unloaded.  In system B, the 45 ft³ of storage would serve the demand before the compressor operates.  The compressor would than have to satisfy the demand and replace the storage before it would unload.  Storage is 45 ft³ at 110 psig or 28.6 lb of air.  The compressor would cycle 62 sec loaded and 35 sec unloaded for a total of 97 sec.  If full load is 22.3 kWh and unloaded is 6.4 kWh, the overall usage is 16.56 kWh or $1.16/hr for electricity at $0.07/kWh.  At 2040 hr/shift/year, the system would cost $2,364/shift/year and would hold a steady point-of-use pressure of 80 psig.

System would fluctuate up to 20 psid and require its 27.5 bhp compressor to run flat out at 22.3 kWh or $1.56/hr for a total of $3,184/shift/year or 34.6% more than System B in electricity for the compressor alone.  That penalty would be at least $1,390 of burdened cost per shift per year or more than $6,000 on a 24-hour basis.

If the design capacity of the compressor in System B is higher than 110 psig at the same flow, the pressure of supply could be increased, thus, increasing the weight flow of supply almost directly proportionate to the rise in pressure.  The horsepower would also increase, but at half the rate of the pressure rise by total percentage.  There would be a substantial improvement in the mass (weight) flow to input power efficiency.  In System B, this increase of efficiency would cause increased unload time and reduced load time.  It is always appropriate to run the compressor a its optimum mass flow to power point once the system is balanced and demand is controlled.

The examples have been created with a demand-supply relationship of 100 cfm at 80 to 100 psig so that it is easy to relate to the effect of these issues in existing systems.  When a full audit is performed on a system, it will find that the cost of air is 1.7 to 3 times the cost of the electricity when the costs of water or air cooling, dryers and filters, accessories, labor (inside and outside), depreciation running and breakdown maintenance, inventory, aging to destruction, insurance, administration, etc. are added.

Without auditing the system from an engineering configuration as well as a financial point of view, it is difficult to get the attention of management.  You may not get management’s support until the next time that you seem to run out of air (or have excess demand) and, in desperation, everyone rushes for a prepackaged engineered power package to throw at the problem.

Reprinted with permission from R. Scot Foss, president of Plant Air Technology, Charlotte, N.C., a company specializing in system auditing and design.  This article is based on his book, "Compressed Air System Solution."  A portion of the proceeds from sales of the book is donated to children’s charities.  To order a copy of the book, please contact Southern Corporation.