Overview

The year is 1989. However, the first automobile rolled off the Ford Twin Cities assembly line in 1925. Since then, the community has grown around it and now crowds the plant boundaries. A paint shop was added in 1984 to help consolidate assembly operations and reduce transportation fuel use (in terms of both pollution and costs). The plant now employs about 700 workers, many of whom live in the surrounding community. The assembly plant produces Ranger and F Series pick-up trucks, approximately 42 per hour, in two 10-hour shifts. The plant closes for maintenance twice each year: for 3 weeks in July and for 2 weeks in late December. Its operations during the remainder of the year depend on truck sales--recently, requiring only half full-capacity.

Painting the truck bodies involves three, sometimes, four operations: first, an undercoat is applied in a dip-tank; second, a PVC/guidecoat/anti-chip layer is applied using electrostatic spray equipment. Next, trucks to be painted in two tones receive their first coat of color; and finally, the last coat of color is applied.

Current Emissions and Controls

1. particulate matter (mostly fine particles of paint), and
2. volatile organic compounds (VOCs) from the solvents used to apply the paint.

Air cleansed of particulates still contains VOCs. The facility emits approximately 1000 tons per year of VOCs. In recent years, this has ranged from 1225 tons in 1986 (the first complete year of operation of the new paint shop) to approximately 750 tons in 1990. The Ford Plant consistently ranks as one of the top ten VOC emitters in the state of Minnesota.

Air containing VOCs from all the operations applying paint is exhausted from a common 20-foot diameter, 136-foot spraybooth stack, with an exhaust gas flow rate of approximately 1.2 million cubic feet per minute(!). This high-volume flow prevents solvents from lingering in the paint shop and is designed to ensure adequate worker safety. The Ford plant thus presents a particular problem for controlling emissions: low VOC concentrations carried in high air volume flow. Reducing emissions of this type is difficult and requires a significant capital investment by the manufacturer. The "best available" technology would be a carbon-wheel absorber, which would reduce VOCs prospectively by one-third to one-half. If installed at the Ford plant, the carbon-wheel would be only the third of its kind in the U.S. Assuming the cost could be spread over the lifetime of the plant (operating at full capacity), this would add about another $20 per vehicle.

Finally, air laden with solvents from the ovens where the paint is dried/baked onto the surface of the truck bodies is incinerated. The incinerators have a VOC reduction efficiencies of about 95% when maintaining a combustion temperature of 1300-1350 degrees F. There are additional lesser sources of VOC emissions at the plant: the final repair and touch-up operations, for example. Unlike particulate matter, which eventually settles back to earth, VOCs in the atmosphere eventually photo-degrade.

Ford has also applied to modify its existing operations by introducing a new "base-coat/clear-coat" (BC/CC) paint. The BC/CC paint is already used by Japanese manufacturers and makes the truck body especially glossy and attractive to customers. The Japanese companies have led the way on this technology; and market studies indicate that it is a significant factor in customer choice and has given the Japanese companies an edge in this sector of the truck market (as well as for other vehicle sales). BC/CC paint, however, uses a different solvent, approximately 80% higher in xylene, suspected by EPA of contributing to irregularities of development in humans. (EPA's current studies on xylene will not be completed for another three years.)

Community Concerns

1. (1) odor, and
2. (2) health effects.

• cancer risks,
• short-term, or acute, effects,
• long-term and chronic effects, and
• developmental effects (on fetuses and newborns).

Studies

• odor complaint evaluation
• community odor survey & mapping
• odor source measurement
• paint chemical composition (particulate matter and solvent)
• (identified toxics and carcinogens)
• emission source measurement
• air dispersion modeling
• wind tunnel analysis
• health risk assessment

Odor

The primary source of odor was identified as the central spraybooth exhaust (main stack). The final repair spray booth and oven exhausts may also be recognizable sources of odor due to their odor emission rates, location and/or low stack height. Further air tunnel modeling revealed that the design of the stack and its position relative to nearby buildings, especially the senior-citizen high-rise, may cause exhaust from the stack to be drawn down to the ground rather than be dispersed in the air. Modifications done under the modeling indicate that raising the stack and rounding its shape would contribute significantly to reducing the downdraft effect.

Health Effects

Cancer. Only two carcinogens were identified as being emitted by the facility: benzene and formaldehyde, each emitted in very small quantities. Cancer risks are typically expressed in terms of probability of contracting cancer during one's entire lifetime. There is no one "safe" or scientifically acceptable level. Much as scientists have adopted a 95% confidence level as statistically "significant," so has the Minnesota Dept. of Health adopted an increased risk of 1 in 100,000 as "tolerable." The predicted lifetime cancer risk from Ford's emisssions--on which MPCA and Ford concur--is well below this level (see table). The risk assessment assumes that people are exposed continuously at the location of highest concentration over their entire life; the likelihood that any individual will fall below this maximum, therfore, may be interpreted as providing a further margin of safety.

[from "Assessment of Potential Health Risks of Paint Spray Facility Emissions, Ford Motor Company Truck Assembly Plant, St. Paul, Minnesota, Technical Report," prepared by EcoSystems Associates [pseudonym], June 1989, on p. 56.]

Short-term, or Acute, Effects. The risk assessment focused primarily on acute effects--that is, those with discernible short-term effects. These are the most noticeable (and, indeed, have been most noticed) by community residents. Both Ford and MPCA studies used a "hazard index" or "hazard quotient" (HQ) approach for expressing the results of risk assessment. In such an approach, one compares ambient concentrations for each toxic against a threshold level at which we now find an adverse effect to be observable. A ratio is computed. Using this measure, if the hazard index is less than 1.0, the concentration of the toxic is deemed acceptable by current standards. In addition, hazard quotients for chemicals potentially having common or combined effects are summed. This combined hazard quotient must also be less than one.

Both Ford's consultant and MPCA assumed that the carbon-wheel absorber would be installed. The results of the two assessments were quite dissimilar, however. Ford's consultant relied on a standard source for acceptable levels of risk for each chemical. MPCA's staff scientist, however, found many such assessments "questionable" in light of more recent studies--studies not yet summarized in a centralized source. The MPCA report, therefore, relied on a patchwork or hybrid of independent assessments, using non-uniform methods.

The first figure in each pair is from the report by Ford's consultant. The second figure (and those for "after stack increase") are from "Ford Motor Company St. Paul Assembly Plant Air Quality Health Risk Assessment," prepared by Fred Adams, Air Toxics Program, Division of Air Quality, Minnesota Pollution Control Agency, Jan. 1990, on p. 9.

Ford's report found the hazard quotient to be below one-half or well-below one-half in all instances. The MPCA report, however, concluded otherwise. It concluded that the potential for respiratory irritation and, to a lesser, extent, eye irritation and headache, nausea and depression of motor or cognitive skills (reaction time or simple mechanical skills) for residents on the upper floors of the nearby high-rise under certain weather conditions, and possibly for people who exercise near the plant "could not be ruled out" (see table on last page). In addition, the potential for emissions to exacerbate a pre-existing condition in susceptible individuals (for example, asthmatics or those with lung disease) could not be ruled out. At the same time, there was no compelling evidence that such health effects would occur regularly, or with the severity indicated by the "conservative" calculations.

Long-term and Chronic Effects and Developmental Effects. Finally, chronic, non-carcinogenic long-term and developmental effects were addressed, but their potential health impact was gauged in each case to be "highly unlikely" and below a level of concern. Reports recognized, however, that there is little information on prolonged exposure for many of the chemicals. MPCA scientists believe, in particular, that the current assessments of health risks underestimate the developmental toxicity of xylene. The developmental hazard quotient for xylene alone (at the elevated receptor) was 0.44 (versus 0.06 on the ground). An increase in the perceived toxicity of xylene, therefore, or a major increase in its emission levels, could generate a possible cause for concern (see also BC/CC paint and nature of uncertainty).

Susceptible Cases. With a modified stack arrangement, the hazard quotient for the upper levels of the high-rise approach 1.0. However, response varies from individual to individual and the actual potential for adverse effects may be affected by age, gender and/or pre-existing health conditions (not reflected in the averaging done for the hazard quotient). Though not much is known, there may as much as a ten-fold difference in susceptibility. For example, children and older people have increased sensitivity to headaches and nausea caused by the VOC emissions: due to their physiology, children experience a more pronounced effect at the same chemical concentration; older persons do not metabolize chemicals as quickly and so eliminate them from their systems more slowly. Also, those who jog or run near the plant, might well experience greater concentrations than were measured elsewhere and while exercising, they tend to inhale larger amounts of air. Effects on asthmatics and on those with pre-existing respiratory disease or cardiovascular conditions are not fully known, here. All cases will be aggravated, of course, by unfavorable weather conditions.

Exposure Assumptions and Uncertainties. The risk assessments provided by Ford and MPCA are just that: assessments. Due to our limited knowledge, one cannot estimate safe levels of chemical exposure very precisely. For example, most (though not all) of our data on health risks has been gathered from short-term animal studies. Chemical concentrations are generally higher, in order to provoke an observable effect (possibly not produced at lower concentrations), while at the same time, cumulative effects from longer-term exposure may be missed. Laboratory animals are genetically similar and healthy before exposure begins; humans often are not. More useful human data is generally hard to collect: for ethical reasons, of course, one cannot set up controlled tests that deliberately expose humans to possibly hazardous chemicals; in data collected from other cases of exposure (say, from those who worked with the solvents before air quality controls were instituted), it is generally difficult to discern the effect of the one chemical in question against a largely unknown background of other possible chemicals. We also do not know enough about how the chemicals interact and whether combinations of certain chemicals reduce or compound their separate individual effects. While effects on the blood and nervous system are relatively well-studied, there is much less known about effects on the immune system, reproduction and development, hearing (loss) and learning.

Where there was uncertainty about possible exposures from the stacks, however, methods "erred" on the side of overestimating exposure. For example, for estimates of long-term exposure, person were assumed to be exposed continuously for 70 years at the point of highest chemical concentration. Clearly, few, if any, persons will experience this actual level of exposure.

Whose Pollution? Finally, measurements of exposure in the area did not separate the effects of the Ford plant from other "background" air pollution that would exist even without the plant--say, from the heavy traffic on Ford Parkway. For example, benzene in the ambient air in a typical urban setting contributes 0.08 to a hazard quotient at ground level, and Ford's emission adds only 0.00006. Ford increases ambient xylenes (the largest proportional increase) by only 15%. Health risks assessments must be based on estimated actual exposure, and so did not isolate those effects contributed just by the emissions from Ford.

Consequences and Alternatives

Because the Ford plant depends directly on the sales of its trucks, the status of the market for Rangers and F-Series pick-ups determines in a close causal connection the plant's work load. Loss of sales means, for those employed at the plant, loss of work and income. In more severe cases, layoffs could be expected. Ford management estimates [let us speculate, for the purposes of this case study] that every $100 increase in the price of the truck results in a 2% decrease in sales. The BC/CC paint, Ford estimates, could result in a 5-12% increase in sales.

Concerned about reducing VOCs and particulate matter in all its paint facilities, Ford has also been researching and developing a new method of applying "high-solid" (i.e., low-solvent) paints. The new method is expected to reduce VOC emissions by a dramatic 70%. Though the method is "ready" for use, R&D scientists are still not confident about introducing the emerging technology on a production scale. The method is costly and will add about $300 per vehicle, at current estimates. Due to the cost, Ford management has not yet committed itself to introducing the new technology. Also, although the high-solid paint involves less solvents, the solvents are also different: the risks associated with the new VOC profile cannot be fully known until the process is installed--and there is always the possibility that the change will generate more odors.

New methods pose special problems when granting permits. Permits are only issued for 5-year periods; however, the technology that is permitted and installed is often "locked-in" for 10-20 years, so that the large capital investment in equipment can be distributed over many years. In some cases--such as the water wall scrubbers, built into the plant itself--pollution control equipment sometimes follows the lifetime of the plant itself--unless the plant is totally rebuilt. Changing the height and configuration (or shape) of a stack, in this context, would be relatively straightforward. This requires modifying existing equipment rather than installing new structures and though still costly, is relatively inexpensive by pollution-control standards. The carbon-wheel absorber and high-solid paint, by contrast, involve major structural changes and substantial capital investment.

Finally, Minnesota has been very strict on pollution control since the 1960s (well after the Ford plant was built). Regulations in nearby South Dakota, however, tend to be much lower. Ford thus has an option to move its plant to an area where the population is less dense and where more lenient pollution regulations promote local industry.