Chapter 6 ProcessCosting

6-2 The United L/N Plant

We never imagined that we'd ever be looking at the type of costing issues at our United L/N Plant that have now become apparent. The plant design was engineered from the beginning as a state-of-the-art production process that would avoid most of the traditional problems. Now we're in the process of taking a second look.

Ken MarvinDirector, Planning and ControlLibbey-Owens-Ford Co.

BACKGROUND

Libbey-Owens-Ford Co. (L-O-F), one of the companies in the Pilkington Group, has been a major producer of glass in theUnited States since the turn of the century. Its newest plant, United L/N, began producing glass products in November 1987.

Located in Kentucky, United L/N is a joint venture between L-O-F and a Japanese company, Nippon Sheet Glass.

Unlike the Rossford Plant, which produces the raw float glass used in its fabrication process, United L/N is a fabrication plant only. The organization of the plant reflects the just-in-time, pull through philosophy. The production process is fully automated, requiring no human intervention from beginning to end. Very high quality and minimal scrap were expected to be the norm.

L-O-F treats the Rossford Plant as a standard cost center, but United L/N is organized as a strategic business unit (SBU).SBUs are evaluated on profit as well as cost control and other goals. The company charges United L/N Rossford's standard manufacturing cost for raw glass transferred between the two plants, whereas transfer prices between SBUs are usually negotiated by their managements.

FABRICATION PROCESS AND PLANT DESIGN

The steps in the fabrication process are essentially the same as at other L-O-F facilities. First, the raw glass goes through pattern cutting where it is trimmed to the basic shape of the lite (window) it will become. Second, the cut pattern is edged. Third, theedged pattern goes through a furnace where it is formed (bent to shape) and tempered. Fourth, the final product is inspected, packed, and shipped.

The principal difference between United L/N's process and the fabrication process of traditional plants lies in the organization of these discrete steps. At the Rossford Plant, for example, pattern cutting and edging, tempering, and packing and shipping are treated as individual cost centers and are physically separated. Each department creates a work-in-process inventory, which is periodically moved to the next stage in the process. The next stage in most cases is located in a different section of the plant.

Also, all processes at Rossford require human participation.

By contrast, United L/N's fabrication process is entirely in-line and automated. Raw glass from Rossford plant and other plants arrives packed on special racks that are designed for United L/N's automated loading process. (The racks are also designed to protect the raw glass from damage between the shipping plant and the receiving area.) The glass has been inspected atthe shipping plant to determine that each piece meets the specifications of the fabrication process.

A forklift operator loads racks of glass at the beginning of one of United L/N's two production lines. From that point on, the entire process is operated through a numerical control computer system. Human intervention occurs only during planned downtime periods when the line undergoes preventive maintenance, when a problem stops the line's progress, and when finished pieces are inspected prior to packing. Glass is continuously pulled through the process, so that ideally there should be no idle workin-process inventory.

A small team of operators monitors the process, performs regular preventive maintenance, changes the computer settings for different lites, and makes unscheduled repairs as needed.

PRODUCT COSTING SYSTEM

The product costing system at United L/N is very straightforward compared to those of less automated L-O-F plants. Overhead costs associated with handling, storing, protecting, and accounting for work-in-process inventories are dramatically lower.

The major components of cost include short-term fixed operating costs, labor costs of the operating teams, and the transfer prices of raw glass from other plants. Because the entire process is automated and in-line, the feed rate is constant across all sub-processes for each individual lite being fabricated. The costing system consequently was designed as one large pool, which is assigned to units based on standard input prices and standard feed rates. The system allows for standard levels of downtime and anticipated yields. Products are not charged for either planned downtime or planned scrap, which were expected to be minimal due to the care with which the plant was designed, engineered, and monitored.

PRODUCTION AND COSTING PROBLEMS

"It wasn't long before we began experiencing problems with our yields," commented Ken Marvin, Planning and Control Director of L-O-F's Original Equipment Business Unit. "At first we thought that the problems would be confined to adjusting and learning about the automated process. We thought that as we gained experience with it we could solve our difficulties without introducing more complicated costing mechanisms."

One of the first difficulties encountered was keeping the furnaces on the two lines working efficiently. Each one was designed to work perfectly when a certain number of glass pieces were being fired, a certain number were on the threshold entering the furnace, and a certain number were leaving it.

"In our traditional plants we stockpile pieces in front of the furnaces so that we can keep them filled to their optimal levels when forming and tempering," Marvin explained. "However, the United L/N lines were designed with no accumulators in front of the furnaces to keep them running efficiently at all times. For any number of reasons there might be gaps in the lines as they enter the furnaces. Partly because of these gaps and the resulting imperfect furnace operations, we have had unacceptably high scrap variances. Of course, scrap decreases the plant's yield."

"From the United L/N point of view the problem with scrap is caused by imperfections in the raw glass rather than by problems with the process. The plant's management therefore believes that the scrap variance should be charged to the shipping plants (including Rossford plant.) Shipping plant managers, on the other hand, believe that the charge-back, even if appropriate(which remains an issue), is much too high. United L/N's costing system costs every piece as if it goes through the entire processrather than dropping out at, for example, pattern cutting or edging stages."

Another factor contributing to furnace inefficiency and ineffectiveness is that a line sometimes goes down unexpectedly because of a problem in pattern cutting or edging. The plant then incurs the opportunity costs associated both with having an empty tempering furnace and with having to reset the furnace after it has been empty during periods of time when it was programmed to be full.

"We have been tracking all sorts of variances trying to get some insights into the effects on costs of the kinks in the process," continued Marvin. "We calculate a combined materials usage and spending variance, a downtime variance, a throughput variance, and a scrap (yield) variance. However, all of them are valued on the basis of costs of the entire production process rather than on the value added to the stages of production where problems occur. Now we're reevaluating the design of our costingsystem at United L/N, especially in light of the ongoing negotiations with Rossford plant and other shipping plants."

REQUIRED:

1. As a member of the Rossford Plant negotiating team, what would be your position regarding the proper treatment of the United L/N scrap variance? As a member of the United L/N team?

2. How could United L/N's management determine the specific causes of defects (for example, bad glass or defective cutting, edging, or tempering operations) in units that are scrapped at the plant? What are the implications for the product costing system?

3. What could be done to solve the problems with furnace inefficiency and ineffectiveness that Ken Marvin discussed? What are the probable effects of your suggestion(s)?