CEC Consultants, Inc.

By Michael Valenti, Associate Editor [July 1995: Vol.117 No. 7]

Ford Motor Company, based in Dearborn, MI, is realizing significant energy savings, reducing capital expenditures, and minimizing wastewater disposal costs by diagnosing and quantifying leaks in its compressed air, steam/condensate, and process water systems by applying algorithms developed by Cleveland-based CEC Consultants, Inc. These algorithms make use of readily available - and often already installed - instruments, such as vortex shedding meters, chart recorders, and data loggers, to compare how much utility use is needed for assembly and manufacturing equipment with how much is being generated.

“For years, I’ve been frustrated seeing plant after plant wasting their money on utilities they’re not using or don’t need,” said John R. Puskar, president of CEC and a mechanical engineer. The reason for this waste, according to Puskar, is that many manufacturing plant operators do not view their utility operations (boilerhouses) as integral to the manufacturing supervision beyond providing the requisite amount of steam, compressed air, or water. Not much thought is given to what happens after steam, air or water leave the boilerhouse.

“The mind-set is ‘if they use it, we’ll just make more,’” Puskar said, which leads to higher fuel or electrical power consumption. Even when plant management and boilerhouse operators recognize the importance of enhancing utility control, their efforts are often frustrated by lack of personnel to diagnose problems or potential savings. “There is tremendous opportunity for applying statistical process control techniques to boilerhouses for system-diagnostics purposes,” said Puskar. “This data can speak volumes about what is happening to the utilities generated once they go to the plant.”

To understand utility use as an overall system and diagnose where leaks and other problems may be, CEC uses readily available boilerhouse data gathered by standard sending devices. The sensing devices. The sensing equipment is located at action points where leaks are likely, such as condensate receivers, key manholes, cooling tower overflows, parts washer overflows, and the air blowoffs on machining systems where solenoids might fail. The sensing devices send their data to chart recorders and data loggers, both of which are usually in place in manufacturing plants or easily installed if not.

The sensors used in the CEC data analysis techniques include orifice plates and vortex shedding meters, which monitor steam and compressed air systems, and transit time (Doppler) meters to monitor water systems. Orifice plates are installed in-line, and as fluids flow through the openings, the sensor measures its pressure drop, which is proportional to flow, over a specified range.

Vortex shedding meters, which are also installed in-line, create vortices as fluid passes over them and measure vortex frequency, which is also proportional to flow. Transit-time meters are externally connected to piping and emit acoustic waves whose disturbances are likewise proportional to flow. Among the manufacturers whose instruments have been used in CEC’s data analysis work are Endress & Hauser Inc. and Polysonics Inc.

Saving Compressed Air
Unlike most compressed air monitoring techniques, which focus on consumption during production time, the CEC technique examines consumption during non-production periods. Production time consumption determines whether there is sufficient compressor captivity, but it doesn’t provide information that could help cut costs, according to Puskar. The load left after production stops, what Puskar termed as “no-load” load, is a more effective perameter for accurately monitoring leaks.

This nonproduction load consists of three components: the real load required to power machinery, leaks too cumbersome to repair, and unspecified problem areas that can be fixed to save energy. Manufacturers should treat their compressed air production as a budget, Puskar said, and balance it. Thus, real load and the amount lost through leaks should be defined with ultimate goals of near zero. Deviations over those two values should be reported as lost potential profit.

Managers of a plant with 24-hour operations, for example, might not want to fix a leak that would take 10 hours of disassembly and downtime to reach. But if they could be shown that a $300 repair job could fix a leak costing $1500 of compressed air annually, they would see it differently.

For example, the Ford Motor Co. stamping plant in Monroe, Mich., has effectively used data analysis techniques to reduce its utility costs. This plant manufactures a variety of automotive parts, including rims and wheels, exhaust systems, and chassis components. Three water-tubed, coal-fired boilers at the Monroe plant can provide a total of up to 200,000 pounds of steam per hour at 160 pounds per square inch gauge (psig). The plant also has a central boilerhouse with nine air compressors. The compressors supply up to 24,645 cubic feet per minute (cfm) of compressed air to 110 psig to all of the large presses, milling machines, cutoff machines, and tube mills that stamp and machine parts made at the Monroe plant.

About three years ago, the Monroe facility consumed 17,500 cfm of compressed air during production hours, and 6000 cfm during nonproduction hours. Subtracting the amount of compressed air consumed during nonproduction periods from the amount consumed during production showed that the plant needed only about 11,000 cfm during production times.

The Monroe plant has a 15-member energy team drawn from its own workforce to analyze data so the plant stays within target thresholds for compressed air use. For example, the team conducted a survey of compressed air users throughout the stamping plant that showed a mere 1000 cfm would be sufficient when production had ceased.

The energy team used the data to detect and repair leaks throughout the stamping plant and, where possible, implement less intensive compressed-air applications. As a result, the Ford stamping plant in Monroe now consumes only 9500 cfm of compressed air, which is produced by a single compressor during production hours. When production ceased during the Easter break in April, the plant used only 1050 cfm. These savings in compressed air use translate into approximately $690,000 in energy savings annually. In addition, a recent production expansion at the Monroe plant did not require new compressors because of the improved control and reduction of compressed air.

Optimizing Steam and Water Use
Analyzing the post-production load works equally well for analyzing steam and condensate systems. For steam, real load and cumbersome leak numbers are relevant only during summer, when the plant does not need to be heated. Here, algorithms could be developed to define and predict expected base loads according to outside temperatures.

At a Ford Motor stamping plant in Walton Hills, Ohio, just outside of Cleveland, workers manufacture the floor pans and body panels for different Ford models, including the Econoline Van, the Thunderbird, and the Cougar. “The Walton Hills plant was using 24,000 pounds of steam per hour for limited periods of the day,” Puskar said. CEC engineers helped Ford cut the cost of summer steam at Walton Hills from $25,000 per month to less than $2000 by eliminating the coal-fired central powerhouse and installing localized natural-gas fired units.

The nonproduction load concept also applies to condensate systems year round by measuring the difference between the steam sent out from the boiler and the condensate that is returned. A significant change in this difference could indicate a problem such as failed condensate return station pumps. “Our firm has seen plants with condensate pouring out of pump stations to sewers for weeks, yet no one in the powerhouse knew. Proper data analysis could prevent this,” Puskar said.

Using data analysis to improve the control of process water consumption is more difficult than it is for air or steam/condensate because of the lack of water data collection devices inside the manufacturing plant. CEC recommends that its clients install domestic water-metering systems that will collect consumption data at key points in the plant.

At the Ford Motor Co. plant in Indianapolis, which manufactures steering components such as rack-and-pinions and worm gears, management wanted to reduce the water consumed by part washing, plating, and cooling towers.

CEC installed Marsh-McBirney Inc. Flow Totes in each of the plant’s 22 key sewer manholes to determine the amount of water being discharged. The Flow Totes consist of a stainless steel ring hung in a manhole. As water flows through the ring, a sensor in the bottom of the unit measures the velocity and depth of the water. An algorithm based on those factors provides the flow rate.

After analyzing the data collected by the Flow Totes, the CEC engineers discovered that the Indianapolis plant was using 300,000 to 400,000 gallons of water per day when production had stopped. This amount was reduced by nearly 50 percent by fixing leaks and improving water-intensive processes, though the plant must consume some water during downtime for certain operations.

Controlling process water through data analysis can also reduce disposal costs, because most spent process water must go to on-site waste pretreatment plants before being sent to municipal waste-treatment plants or discharged, Puskar said.

Ford is the largest single user of the CEC data analysis techniques, employing them at 13 manufacturing plants to date, but CEC’s techniques are also applicable to other industrial facilities that rely on central utility systems. For example, CEC engineers have worked with Miles Laboratories, in Elkhart, Ind., to monitor process water used to manufacture pharmaceuticals.

The CEC data-analysis techniques could also possibly be used to monitor electrical utilities. “Waste in the power generation is typically included in the overhead cost, without any effort to correct it,” Puskar said. “Using data analysis to reduce steam consumption by eliminating waste could make power plants more efficient, extending fuel reserve while reducing emissions.”