With Production Operators plant in Evanston, WY under contract to Chevron to produce and compress nitrogen gas to 3,000 psig pressure, we were relying on the air separation plant purchased from Kobe Steel in Japan. In early tests, the plant had failed dismally to produce anywhere near the contracted nitrogen volume, had failed to remain on-stream for more than 10- 12 days and now, with Chevron applying substantial pressure on us, we had just had a big explosion and had been compelled to completely shut down.

Conveying the enormity of this to Kobe and their marketing representatives, the Mitsui Corporation was impossible through their on-site start-up engineer who had practically collapsed just looking at the result of the explosion, so it had been decided that Al Domeshek and I would go to Tokyo and Kobe to explain the situation to both parties.

When Al called and told me about the explosion, I had immediately gone to the plant and was able to fairly quickly identify what I believe had happened by examining the recorder charts and the operation log together with the physical evidence. The explosion had blown a hole at the bottom of the cold box near the expansion turbine which produced the refrigeration for the plant. There had been about six inches of snow on the ground but the area around the turbine was dry and charred showing that there was a brief flash of fire. To have an explosion, there must be a fuel, an oxidizer and an initiator so we looked for these three factors. The fuel was clearly the oil-soaked ground around the turbine which we found had had some lubrication problems and, apparently an oil spill. The presence of an initiator, unless it is very obvious, is often hard to determine but if oxygen rather than air is involved, the immediate suspicion falls on an atmospheric electrostatic charge, particularly in the cold, dry air at the roughly 8,500 foot elevation of the plant.

This led us to look for an oxygen or oxygen-rich source. When we examined inside the cold box we were unable to find a rupture or a potential leak site in either of the two distillation columns but we noticed that there was no air-tight foam insulation on the critical surfaces near the top of both columns that we were accustomed to seeing in all air plants. In the operation of all cryogenic plants, if temperatures of the process go below the condensation temperature of atmospheric air it is necessary to insulate the surfaces with an air-tight foam or other insulant to prevent the air condensing and accumulating somewhere or to stop it running down the metal surfaces. In the latter case, there could be a distillation effect with the liquid becoming more oxygen-rich as it falls and partially vaporizes.

In air separation plants, the normal temperatures at the top of the low pressure column and even in certain circumstances, at the top of the high pressure column, are below the saturation temperature of air and condensation can occur, giving exactly the same appearance as steam condensing on cooler surfaces. Our combined opinion was that this is exactly what had been happening and I was even more sure when I re-examined the pressure recorder chart for the high pressure column, I discovered a severe pressure drop about fifteen minutes before the explosion. Undoubtedly the extra condensation from the second column created enough oxygen-rich liquid to run out of the bottom of the coldbox and onto the oil-soaked ground.

On January 10^th, 1981, Al and I, together with Takuji Sakuraba, the Mitsui liaison engineer we had always worked with, flew to Tokyo on Japan Air Lines via Los Angeles, with the second stage taking 11 ½ hours. Tak was Mitsui’s resident engineer in their Houston office with whom we had been handling all Kobe matters and he had made all of the reservations for the trip and subsequent meetings. He was quite fluent in English and he had told us we should both bring two bottles of Chivas Regis scotch as gifts, the recipients of which he would choose for us and handle confidentially.   Knowing the Japanese business culture as he did, he stressed that this was important and probably expected. We had travelled on a Sunday and knowing the jet lag well, Tak had planned our Monday meeting with the Mitsui management for the afternoon. We checked in to the Tokyo Park Hotel, had tea and a snack and went to bed.

Our meeting was with a group of non-technical senior managers and was quite formal with much bowing with the introductions.   (I had warned Al to not repeat his anti-bowing speech to the Japanese plant engineer when he first arrived at the control room in Evanston, when he colorfully told him that we don’t bow in this country.) To our surprise, all they seemed to want to talk about was the explosion and what POI was doing about it and seemed to be oblivious about the performance failure of the Kobe plant as if the problem never existed. In general, we decided that they wanted no blame to fall on Mitsui and if we had a problem it was to be discussed with Kobe. When seven of these managers showed up at our dinner table for ten in an obviously expensive Chinese restaurant, we realized we were being used as an excuse to provide them with a company paid dinner. The conversation was almost totally among themselves in Japanese and Takuji and his immediate boss were the only ones we spoke with. The table was a giant Lazy Susan carousel type and large quantities of food that Al and I had never seen before revolved around in front of us.

We sampled the ones we recognized and I experimented a little but Al stuck to meat, fish and veg.   It was a very interesting evening for our first in Japan and I think we learned a little about the culture of “Japanese Denial” we had been told about.

In the morning we left Tokyo for Kobe on the Shinkansen or bullet train which took us at speeds up to 180 mph the 265 miles in well less than two hours. On the way, we had a great view of snow-topped Mount Fuji and the only other memorable moment was when we met the Shinkansen going at the same speed in the opposite direction.

In Kobe, they had scheduled a morning
meeting for each of three days with their engineering department. The three of us were always on one long side of the table and a phalanx of people on the other, clearly seated strictly according to rank. We first told them our explanation of how we thought the explosion was caused and this filled the next hour with Japanese-only discussion, shouting and head-shaking. We had difficulty deciding if they had understood us and had no idea whether they had accepted our explanation or not. One group left the meeting and we never saw them again. The more senior members spoke some English and asked very elementary questions demonstrating little engineering knowledge and when we broached the question of failed performance, they immediately said that that would be discussed the next day. We tried to get them to commit to fixing the plant damage and to insulate the columns knowing that we would later be asking for far more stringent remedies. That afternoon, Tak took us on a tour of Kobe but our discussions were solely on the obvious difficult task we would have the next day demanding substantial changes to their plant. How it worked out is another memory for another day.