Why is it people with the least knowledge or experience in an operation can make decisions that supersede the accumulated knowledge of the entire staff?
Our operation employs wet-grind ballmills to pulverize sand to a usable size for our process. The ballmills are fed by belts. The feed rate is controlled by belt scales that give us a tons-per-hour calculation and input to the PLC to control the belt speed as the sand varies in pounds-per-inch on the belt. This is an almost foolproof system.
Enter the fool. He's a smart enough guy in his field. Unfortunately, his field is pulp and paper, and he has a degree in chemical engineering. He graduated from college with his BS this year, and this is his first real job. His first question on seeing the ballmills was, "What's in the barrels?"
If you just cringed, welcome to my world, brother.
Our conscientious young engineer takes his job seriously. Unfortunately, he also takes himself seriously. He doesn't take some uneducated old bum like me seriously, because I don't have a degree. My 20 years of experience and the fact that I'm responsible for programming the PLCs that he can't even look at doesn't faze him. He's got a college degree; I'm illiterate.
Now for the problem. He didn't trust the calibration of the belt scales, so he decided to check it himself. He and the leadman from his department gathered their gear and went to work. First, he got a hand tachometer and measured the belt speed. He was apparently puzzled by the speed changing. After all, the sand wasn't changing much. He finally decided to take two measurements and average them.
Next he stopped the belt through the simple expedient of opening the disconnect. He's never heard of "turn it off first." Our VFDs are fairly tough, though, and he didn't destroy anything. Then he brought out his trusty Stanley measuring tape and marked off a four-foot long section of the belt and drew a line across the sand with his pocketknife. This was going to be his "sample." He and the leadman scraped the sand from the belt into five-gallon buckets, then weighed them. The precision weighing instrument they used was a brand-new $39.95 bathroom scale from Wal-Mart.
After weighing the sand and subtracting the weight of the buckets, the fount of all wisdom calculated the sand scale was out of calibration by nearly two tons per hour. Horrors! He was incensed! Infuriated!
He immediately radioed me to come to the ballmill. The scale was out of calibration and had to be fixed immediately. So I came. I saw. I laughed till I blacked out. Then the maintenance manager du jour showed up and straightened me out. I'm just an electrician (well, plant programmer, senior electrician, and the only certified instrument tech in the plant), and I'm to do what the engineer tells me to do. So, in accordance with my instructions from both the engineer and the maintenance manager, I adjusted the calibration of a $10,000 belt scale to give a reading that matched the accuracy of a $40 bathroom scale. What the hell, I'm hourly. It doesn't affect my bonus (don't get one).
The result of the scale calibration was evident within three days--a complete catastrophe. Calculated sand usage was way off from recorded sand usage. The quality of the sand coming out of Ballmill 1 was horrible. Something had to be done. So they blamed me. I had obviously done something wrong when I recalibrated the scale.
Our young engineer once again brought out his trusty bathroom scale and buckets and went to work. He measured carefully. He weighed the buckets and got readings to within 1 lb. on the scale. And once again, his calculations didn't match what the belt scale said. So once again, that scale had to be recalibrated. This time he stood beside me with the book out to see what I was doing wrong. Once he was satisfied I had done it right this time, we closed up the scale and he put his lock on it to make sure nobody messed with it.
Remarkably enough, the results got even worse. Now it was so bad the slop coming out of the mill wasn't usable in our process. Since I obviously didn't have the skill to calibrate the scale properly, they called for a tech from the company to come out.
This guy is my hero. He showed up the next day and got the whole rundown from the engineer. It was mostly me he was running down, but what the heck? When he finished, the tech looked at him for about 10 seconds,then started laughing. He packed up his gear and went back out to his truck to call his boss. After about 10 minutes, he came back in.
You probably guessed what happened. We had, by improperly calibrating the scale, voided the five-year warranty. He had us run the belt empty, then run through the entire calibration using his test-weight chain. Once the scale was calibrated we compared the numbers that had been recorded during the initial calibration to what we had then and found that the calibration had drifted 50 counts,in a 20,000 count range. The last number I had put in to make the engineer happy had been almost 8,000 counts off.
By now the entire upper office was watching to see what was happening. Our production and profitability were on the line. And the engineer was busy explaining to the plant manager why he thought buckets and a bathroom scale were more accurate than the scale that was made for the situation.
Needless to say, the bathroom scale went to someone's home. The bulletproof young engineer got assigned to a less critical position. I think he is mixing paint in California. And the belt scales don't get messed with except once a year when we calibrate them with the manufacturer's test weight.
Loren Jones is electrical leadman, senior instrument tech and plant programmer at James Hardie Building Products, Plant City, Fla.
Tales From the Front is selected from reader submissions. Send 800-900 words that you as a process control end user, system integrator, or consultant would like to read in CONTROL (a problem you have solved; a sage observation about plants, people, and process control; a technical detail you want to share with 70,000 others"the subject is up to you) to email@example.com. Upon publication, we'll pay $500.