In 1982, I was working for GEC Engineering Services, in Germiston. I reported to Pierre Ballot, who was an incredibly good engineer. My job was to test and commission electrical equipment that had been made by other GEC organisations.

I don’t think I was very good at my job. The problem was that the electrical equipment was fairly complex and dangerous and could injure you if it had a fault. This scared me. A project came up at the mining town of Oranjemund. This is a mining town at the mouth of the Orange river. For years, the town and mine had power supplied by a diesel power station rated at 18 MW, which supplied power at medium voltage. Then the power supply authority put in a brand-new 220 kV power line to supply the town and the diesel station was decommissioned.

After a year, an aircraft flew through the power line, and Poof! The diesel station ran again but not all of it worked. GEC was given the task of upgrading the station. At GEC, Win Lees redesigned all the control desks and relay protection panels and I tested them. They were shipped off to Oranjemund and installed, using a cable and connection schedule that I drew up. Then, after a few weeks, I was sent off to test and commission the installation. I loaded up a 1300 Cortina sedan and rolled off to Oranjemund, going through Potchefstroom, Vryberg, Kuruman, Upington, Karasberg, Port Nolloth and Alexander Bay. A long way. North of Port Nolloth, the road was made of compacted salt and the rain (a rare occurrence) made it slippery. Halfway to Oranjemund, I came across a car on its roof. The occupants were unhurt. I undertook to tell the people at Oranjemund.

I arrived and started work the following day. There were no cellphones and no computers in those days, so I used the public phone to tell my wife I was safe. This involved sitting in a waiting room after I had booked a call. Took about two hours. The electrical panels we had sent were at the power station, all connected up but the diesel control panels for each alternator had been taken out and scrapped. This was a big unanticipated problem. So, the following day, after doing tests, I sat in the waiting room and finally phoned Lees. I said I could rebuild the diesel panels from the old ones but needed speed-sensing switches. He said he would send some. I spent the week doing tests and rebuilding the diesel panels. It took forever. Every part was damaged or missing. Finally, the speed-sensing switches arrived and we could start up and synchronise the diesel generators one by one. One generator we just could not load without it tripping out. I now know it must have had a problem with the automatic voltage regulator. We tried everything.

After two weeks, I wrote a report to Ballot, detailing progress or lack thereof to date. A big issue was getting the automating starting sequence system to work. It was driven by a series of cam switches which started various auxiliary drives in sequence. Sometimes it would do the job perfectly. Sometimes it would start the drives in the wrong order. The cam switch was fitted with two capacitors and, somehow, these could be used to set a sequence. I went back to the sitting room to phone Lees. It was full. So I waited until 20:00 that night and called after a long wait. He listened to my problem. Then he said that he was on his third brandy and Coke and he hadn’t a clue. Now, 38 years later, I remember the whole experience. I finally solved the cam problem. The whole job took two months. If I had had a cellphone, the Internet, a programable logic controller, an automatic diesel control and synchronising system . . . it would have been quick. But I wouldn’t have learned so much. I’m glad of that.