Process

The facility’s original compressed air system consisted of three compressors that were each controlled separately without a system controller. Originally, during production, the system demand was around 1100 cubic feet per minute (CFM) and on weekends, when there was no production, the demand dropped to 800 CFM, which indicated a significant volume of artificial demand. Table 1 summarizes the specifications for each of the original compressors.

Table 1: Existing Compressor Specifications

The plant’s compressed air is piped into the plant and separated into four 4-inch branches for different end uses. Header 1 feeds Resin Cooking; header 2 feeds Resin Packing; header 3 supplies air to fire risers and HVAC; and header 4 feeds Grinding and Compounding. An ultrasonic air flow meter was rented to determine airflow throughout the plant. Table 2 summarizes the airflow as measured to each of the headers.

Table 2: Airflow Measurements

PLENCO’s goals initially focused on implementing compressed air leak detection and repair practices. They quickly expanded their approach to include data collection and analysis to identify opportunities for overall system improvement. PLENCO completed leak repairs and then installed flow meters and valves to isolate additional compressed air leaks and measure compressed air demand during shutdown periods. The engineering team worked with maintenance to measure compressed air in parts of the facility where production was shut down. For example, header 4, which serves Grinding and Compounding, measured a flow rate of 450 cubic feet per minute (CFM) during non-production hours. By isolating airflow to the Grinding and Compounding Departments during the weekends, they can run the plant using only the VFD compressor, Compressor 1.

Additional compressed air savings during the weekday period were realized after PLENCO modified operating hours for the Compounding department to two 8-hour shifts from the normal three 8-hour shifts. Engineering added a second set of isolation valves to header 4 between the Grinding and Compounding departments. The valves further reduced compressed air demand, but since the original compressor #2 had manual shutoff and modulation controls, it was not able to run effectively at partial load. Originally, compressor #2 airflow was controlled by inlet modulation, but inlet modulation is not an efficient way to run a compressor at part load. Based on the capacity adjustments that were made, the engineering team saw an opportunity to implement a base-trim control strategy with the existing VFD compressor.

According to the principles outlined in the Compressed Air Best Practices bulletin, Control Strategies for Efficiently Operating Multiple VFD Compressors, a VFD compressor should be 30% larger than the base-loaded compressor to avoid a control gap or poor compressor control.  Using this strategy, engineering decided to replace the old compressor #2 (200 hp) with a smaller 100 hp compressor. The new compressor #2 would not have modulation but would run load/unload, a much more efficient control strategy when coupled with a separate VFD compressor. It also can stop automatically if it runs unloaded for five minutes. The new compressor #2 was set up to maintain pressure between 87 and 102 psi. Since the capacity of the new compressor was 501 CFM, it could not meet demand on its own during peak production and would run fully loaded. The decision was made to use this unit as the base compressor within the new control strategy. To meet the remaining demand, compressor #1, with the VFD, would then operate as a trim compressor to meet the varying operational loads.

The minimum flow for compressor #1 is 190 CFM, and the maximum flow for the new compressor 2 is 501 CFM. On backshifts and weekends, the compressed air demand would drop below 691 CFM and the pressure would rise to 102 psi until compressor #2 is unloaded and shut down automatically. Compressor #1 would then meet the varying demand by itself. When demand increases beyond the capacity of compressor #1, the pressure drops to the minimum setpoint, 87 psi, and Compressor #2 starts and loads automatically.

Measuring Success

PHASE I

Based on the flowmeter installed outside the compressor room, PLENCO was able to reduce weekend compressed air consumption by 62%, from a daily average of 1,328,000 SCF to 510,000 SCF after installing the first header isolation valve. The Electrical Department also installed a meter on the original Compressor #2 after the completion of the first header isolation valve set. During weekend periods, Compressor #2 would draw 182 kW with the header isolation valve set open, causing 454,200 kWh of energy consumption per year. The electricians then installed a meter to Compressor #1 with the header isolation valve closed. The average load dropped to 60 kW resulting in energy consumption of 148,700 kWh per year during weekend operation, a reduction of 305,500 kWh annually. This resulted in $26,200 worth of annual energy cost savings, exceeding Engineering’s initial estimate for the project. 

PHASE II

The total cost to purchase and install the new compressor #2 was $43,100. The local utility program provided an incentive for $11,400 to replace the compressor and the replacement achieved 243,000 kWh of annual savings. The reduction in electrical cost provided savings of $20,900 per year and produced a simple payback of 1.5 years. This was the result of running the existing VFD compressor more routinely and effectively. Furthermore, actual system pressure was reduced slightly with the smaller compressor #2. This reduced artificial demand.

PHASE III

After the installation of the new compressor #2, the second header valve set for Compounding ran better and production is now able to close and open the header valves without any action by the compressor operator on both weeknights and weekends. Production started closing the second header isolation valve routinely. This reduced the total airflow on the backshift and assuming this load would be on Compressor #1 results in a power reduction of 41.4 kW. Assuming this valve is closed at least 200 workdays for 6.5 hours results in a reduction of 53,800 kWh and $4,630 per year in energy cost savings. This improvement paid for the original cost of the second isolation valve in less than one year. 

Outcomes

Actual air consumption has reduced significantly since the start of all three phases. The average daily airflow from the compressor room decreased from 1,479,000 SCF in 2018 to 772,000 SCF after all three phases were completed. The success of the header isolation valves has led PLENCO’s engineering team to start installing isolation valves on individual compounding bays. When a bay is not running, production turns a switch to “off” that closes the actuated compressed air valves to isolate all compressed air from the bay. This will eliminate all air leaks from each isolated bay. At the bay level, this can save up to 25 CFM. Maintenance has already completed this change on four compounding bays and the plan is to continue isolating even more bays.