Karen M. Luken and Tim Smith

Not so long ago, collection managers faced severely limited funding options for refuse collection programs. As municipal budgets shrank and the volume of trash escalated, volume-based rate structures gained popularity.

"Pay-as-you-throw" programs offer an incentive for residents to divert recyclable materials from the waste stream while teaching them to avoid products with excess packaging.

Since disposal fees are often based on weight, volume-based rates are truly equitable if the refuse density inside the containers is constant throughout the customer base. Haulers stand to lose money when residents pack their containers solid to take advantage of rates based on the average density of uncompacted garbage.

Because of rate inequities, innovative haulers are trying to find a way to bill customers by the weight of the refuse instead of the volume.

As early as 1991, one Minnesota city began to explore weight-based billing. In addition to finding an equitable way to bill residents for trash collection, the city of Farmington wanted to:

* Increase waste diversion by 10 to 20 percent;

* Reduce collection costs; and

* Increase the quantity of recyclables at the sorting center to improve efficiency.

With these goals in mind, city officials, geared with scales on their collection vehicles and computers in the cabs, ventured into the unknown.

If the city's program had become fully operational, it would have billed customers a service fee of approximately $22.00 per month to cover equipment, overhead and labor costs plus 2 to 3 cents per pound for disposal.

Although the program did not succeed and was never fully implemented, its results could be beneficial to other communities, said the EPA-sponsored Municipal Solid Waste Innovative Technology Evaluation (MITE) Program. Shortly after the pilot, MITE hired SCS Engineers to determine where the program failed.

With about 7,000 residents, Farmington is two miles southwest of Minneapolis/St. Paul, and is primarily flat and mostly residential. The project's service area included approximately 2,200 residential and 135 commercial customers.

The pilot used two fully-automated collection vehicles with mechanical arms. The arms, which were operated inside the cab, gripped each 30- to 100-gallon container, lifted it into the truck, emptied its contents and replaced it at the curb. The dump cycle took about 12 seconds.

The city installed load cells on its collection vehicles, which measured the container's weight through the strain on the mechanical arm. In order to legally bill customers, the National Institute for Standards and Testing (NIST) requires a static-weigh system in which the mechanical arm stops the lifting motion for a few seconds to allow the sensing device to measure the weight. Each container must be weighed full and empty; a standard tare weight cannot be used. Approximately 8 seconds are added to the dump cycle when the arm stops to weigh the containers.

To match the container's weight with the customer, the city tested two tracking systems. The first used a printed route sheet with bar codes as its address database. The route sheet was fed under a bar code reader inside the cab to generate address information. The computer recorded the address and weight.

In the other system, radio frequency identification (RFID) transponder chips were mounted to each customer's container. When the cart was emptied, a receiver detected the transponder's identification code and sent the information to the on-board computer, which decoded the identification number into a street address and recorded it with the weight information.

The Lessons Learned As one of the first communities to test a pay-as-you-throw program, the city had very little information on which to base its decisions. The program's obstacles included:

* Durability Of Load Cells. The load cell manufacturer was not warned about the forces encountered by the mechanical arm (lifting, accelerating, changing direction and emptying and replacing containers). As a result, the load cells, which were designed for light manufacturing and have never been used for refuse vehicles, could not withstand the applications. Each repair to the load cells was charged to the city.

The city could have formed a partnership with a manufacturer who could use the city as a research and development test site. Possibly the equipment could have been supplied at little or no money.

Many manufacturers have developed scales for the waste industry. In addition, had the city issued performance specifications, repair responsibility would have shifted to the manufacturer.

When purchasing load cells, remember that performance specifications dictate what the equipment is required to do; design specifications, however, dictate how the equipment should be made.

* Weighing Equipment Inaccuracies. "Meeting Class III standards is the most significant obstacle to this technology's success," said Robert Williamson, the city's solid waste coordinator.

Class III standards provide tolerance limits for scale accuracy for commercial applications (billing customers). A scale with onepound increments must be accurate within one pound for the first 500 pounds; within two pounds for 501 to 2,000 pounds; and within three pounds for 2,001 to 4,000 pounds. These standards must be met when measuring container gross and tare weight.

For streets on hills or sharply crowned roads, Class III standards requires an out-of-level switch which disables the system when the scale is more than 3 degrees out of level. Also, a shift test must be conducted where the container load's center of gravity is shifted vertically or horizontally. This simulates lifting a container with materials which would settle or move when lifted.

Farmington was never able to consistently achieve Class III standards for accuracy. Most of the time, the project was accurate within approximately 2 pounds. The inaccuracies have been attributed to the automated arm, which provided an unstable platform for weighing. In addition, truck vibrations and crew movement were amplified by the mechanical arm. Lastly, environmental factors such as wind had a significant impact on the scale's accuracy.

Technology meeting Class III standards for semi-automated collection vehicles is emerging, according to scale manufacturers. The load cells will be mounted directly between the cart tipper and the truck body; the mechanical arm, which amplifies truck movement, is not necessary.

Instead of weighing the container as it's lifted, the refuse truck can be weighed after each container is emptied. On-board systems to monitor the total weight of the load include a scale mounted between the frame and truck body. These scales, which generally use increments of approximately 25 to 50 pounds, also can prevent overweight vehicle fines.

Other weight mechanisms are be-ing developed. One company, for example, weighs the loads by using sonar to measure the strain on the truck body after each individual container is emptied. This system provides a gross weight of the refuse in the truck, with the tare weight being the gross weight measured for the previous container.

Scale systems achieving Class III accuracy for semi-automated trucks also have been developed. Designed for rear or side loaders, the load cells are mounted between the cart tipper and truck body. The relative stability of the semi-automated truck allows the load cells to achieve the required accuracy.

* Container Identification System. Although the bar code reader system was reliable, it required drivers to follow the exact sequence on the printed route sheet. Since the route sheet was continuously fed under the reader, drivers couldn't alter the route due to traffic conditions, repairs or special pickups.

The RFID system suffered from equipment failure. The full duplex transponder system required the antenna to be mounted within 18 inches of the transponder chip. When mounted on the mechanical arm, the antenna knocked against the truck body and was damaged.

A local welder designed a separate arm for the antenna to bring it closer to the container during the dump cycle. The second mechanical arm's antenna was not subject to the forces of the container-lifting arm.

Without major modifications to the vehicle, the RFID system could not accurately read the cart-mounted transponders. The antenna's large size and short read range hindered this system's abilities.

A more flexible RFID system that uses FM transmission, half-duplex technology for its transponder identification mechanism has since been developed. The antenna sends a pulse of energy to the transponder, which uses the energy to charge a capacity contained in the chip. The energy lasts approximately 50 milliseconds before the antenna listens for a response. The system's read range is up to four feet so that the flexible antenna can be mounted on or near the truck's hopper, where the carts are identified as they are emptied. These systems are currently being used in refuse collection across the country.

Industry experts agree that accurate weighing systems for automated vehicles will be available one day. Equipment manufacturers are daunted by Class III accuracy requirements for automated trucks but systems for semi-automated vehicles are expected to become reality soon.

It's not easy to know which technology will benefit a community's collection program. While innovative pilot programs may not directly benefit a specific community, the successes - and hardships - will prove useful in developing collection technology for the future.