* * *
At the Pine Bend refinery, Koch was allowed to expel an average of 8.3 kilograms of chromium every day and 714 kilograms of ammonia. That was the letter of the law. But Faragher also wanted to abide by the intent of clean-water laws. Obviously, the intent of the law was to keep large levels of ammonia out of the nation’s waterways. That’s why the limit of 714 kilograms was set. But setting the limit at 714 kilograms did not mean that regulators wanted Koch to pump 714 kilograms per day into the Mississippi River. The state had set a maximum level of pollution, but the goal was to be under that level. The intent of the law was to encourage Koch to pollute as little as possible.
With that in mind, Faragher designed a water treatment plan that kept ammonia and other toxins at very low levels. When it came to measuring pollution, everyone in her business used the terminology of “parts per million” to figure out how much pollution was leaving the pipes with each gallon of water and how close the company was to hitting its limits.
If the mandatory limit was forty parts per million, Faragher liked to keep the flow at about twenty parts per million. This was a habit that she’d learned at the paper mills. Doing so gave the company a large buffer. Treating water was an inexact science, and there was bound to be unexpected spikes in the level of pollution. Keeping the normal pollution rate low helped the company avoid busting its permit levels in case of an emergency. But maybe more importantly, running at a low rate of pollution helped the company meet the intent of the law. The point was to keep waterways and air as clean as possible.
One day Steve David came to Faragher’s office for a talk about the Koch method of wastewater treatment. He drew a large graph on a white board in her office. There was a straight line that ran across the graph from left to right: that was the permit level. Ammonia levels could not exceed that level. Below this line, David drew a squiggly line to represent Koch’s actual ammonia emissions (the line was squiggly to represent the natural variation in daily ammonia levels). There was a big gap between the squiggly line and the straight line. This represented that Faragher was emitting far less ammonia than the permit allowed.
David pointed to the line representing low levels of ammonia emissions and said, “You don’t have to run here,” Faragher later recalled.
Instead, David drew a new squiggly line that ran just below the permit level. David told Faragher that she could run pollution levels there, just below the maximum level. The goal was to keep the ammonia output levels stable. If they avoided big spikes in ammonia output, they could operate just below the level permitted under the law.
Faragher listened intently. She understood what he was saying. Running pollution levels just below the permitted level might seem good for the plant—it was a way to avoid expensive treatment procedures. It was also a way to make sure that the plant was able to run continuously at high volumes. In theory, this plan would work perfectly. There was nothing about Koch’s approach that was illegal. Koch could keep its pollution levels just below the legal limit and still operate within full compliance of the law. And the Koch engineers prided themselves on being smart and running efficiently. They were just the kind of people who could keep pollution levels right within the narrow band that they aimed for.
But Faragher was uneasy with this method. It counted on things going just right inside the oil refinery. That wasn’t how life really worked.
* * *
Things started going wrong around June 1, 1996.
The problem started inside a machine called the sour water stripper, which played a critical role in cutting down on the ammonia emissions that were pumped into the wastewater treatment plant. For some reason, one of the sour water strippers started to malfunction. Only later would it be discovered that a series of trays inside the stripper had built up a layer of residue called “scaling” that made the trays far less efficient. Unfortunately, no one at the plant was aware that this had happened. The trays were buried deep inside the machine, and they could not be seen unless the machine was turned off and disassembled. Doing so would require a partial outage at the refinery. Production would be interrupted. Output would fall. Sales would be hurt. The sour water strippers were allowed to keep running.
Large levels of ammonia started flowing into the wastewater plant. Only so much of the ammonia could be removed by the nitrification process. The ammonia loads were overtaking the microorganisms that were supposed to eat them. As a result, heavy doses of ammonia were sent from the wastewater plant out to the polishing ponds and, ultimately, into the river. Doing this for too long would violate Koch’s operating permit. The fine for doing so would have only been about $30,000—pocket change for Koch. But it wasn’t the fine that was important. If the high ammonia levels still continued, then the legal troubles would escalate. The entire operation could be endangered.
Brian Roos discussed this problem with the other people in operations, like Todd Aalto in the wastewater treatment plant. Roos also talked with Aalto’s boss, a woman named Ruth Estes. These discussions often occurred when Heather Faragher was not around. The operations people had to figure out how to handle the high ammonia levels. Eventually they settled on a rather elegant solution. It wouldn’t bust the permit levels. But it wouldn’t require the refinery to shut down, either.
* * *
From the control room, Todd Alato could pull back on the amount of water that was flushed into the polishing ponds and divert it into a series of large detention ponds on the far end of the refinery. These ponds were enormous: one held twenty-two million gallons of water, and another held twelve million. The ponds were a kind of catchall basin for runoff from the plant, and it was not entirely clear just what was inside of them. The ponds took runoff from the oily water sewer and other pipes within the refinery—the cracked and leaky system where employees dumped naphtha, xylene, and other chemicals. A test by state regulators would later show that soil near one of these detention ponds was contaminated with mercury, chromium, zinc, and other pollutants.
In June of 1996, operators like Aalto started sending millions of gallons of water that was heavily polluted with ammonia into these detention ponds. The technique was known as “stacking” the water, and it had the immediate effect of helping Koch Industries. Because the ammonia-laden water was being stored in these detention ponds, it was not being sent out to the river, where it would count against Koch’s permit levels. Engineers like Faragher were testing for pollution in the polishing ponds, not the detention ponds.
Unfortunately, over the ensuing months, stacking the water began to present its own problems. Water levels at the detention ponds rose steadily. Soon, the water was creeping dangerously close to the tops of levees that surrounded the ponds. If there was a heavy rainstorm, it could potentially cause the detention ponds to overflow, sending a stream of pollution into nearby farmland and wetlands.
Roos assigned engineers to figure out just what was causing the high ammonia levels, but they failed to do so. For a number of technical reasons, the engineers did not suspect that the problem was the trays inside the sour water stripper—usually those trays were effective for many years, and the trays in place then were not very old. The sour water strippers were not shut down and disassembled for rigorous inspection.