Traditionally, the industrial Purchasing function focuses on procuring needed parts, materials, and services. At times, those decisions made in Purchasing have direct negative impact on equipment performance and reliability. It’s easy to connect the dots when parts are substituted and reliability suffers, but less so when production raw materials are substituted.

As RAM pros, we often get concerned when Purchasing personnel seek alternate sources and “like kind” specifications of spare parts and components for plant equipment. Then, again, those individuals are charged with saving money, and spare parts can be a big expense for the business. So, Purchasing’s perspective is frequently price, lead time, and availability. We, on the other hand, are always concerned about the performance and reliability of anything installed in or on OUR equipment (and should be about anything running through it). These concerns are unknowns that buyers can’t address.

We can evaluate performance and reliability data of alternatively sourced parts over time if we must. And we can conduct acceptance testing if we have the capabilities and time. Those activities, however, are also a cost to the business, especially if the new parts cause delays or, worse, extensive or catastrophic downtime. The same holds true with substitutions of raw materials for production. In fact, the impact of production raw materials on equipment reliability can be enormous. Alas, this issue is often overlooked. Let’s examine it through the following real-world example.

RELIABILITY AND PRODUCTION RAW MATERIALS
In my early years, I focused on identifying and addressing reliability issues with traditional reliability-improvement tools and remedies, mostly from a maintenance perspective. When raw materials going into a manufacturing process led to functional failures, we often pointed the finger at purchasing and their changes in sourcing and specifications. Production would fuss, but, in the end, lived with Purchasing’s choices. Reliability improvement then changed to a Production issue. Those of from the Reliability and Maintenance sides of the house had a different view: We knew more could be done.

We assembled a small team that included top-skilled operators, a Production supervisor, and a Quality manager. The piece of equipment being studied was a bottleneck in the production process. It had to perform as intended to keep production flowing through the downstream processes that resulted in finished goods delivered to the customer. In this case, we were called in to look at a rash of jams and higher-than-normal defects and setup scrap (waste) materials. The operators could baby the machine to handle the materials, but the process was slower. At first, they blamed the tooling. While there were some minor tooling issues, they were just contributing factors, not root causes of the problem in processing the material.

Had the operators not noticed a different inventory ticket on the skid of raw materials, we might never have know about  Purchasing’s decision to change suppliers. You can imagine the fingers all pointed to the Purchasing Department as the culprit. The Purchasing manager, though, went to the mat in assuring us the new material was identical to the former supplier’s and that it also represented significant cost savings and shorter lead times for delivery to the plant.

At this point, our team of investigators decided to follow the new material from the production process to the downstream assembly operation to determine if any differences had been noticed. And yes, production had seen a big difference in the new material, not only in the way it ran in the assembly operation, but also in highly variable and inaccurate skid counts

LET THE TESTING BEGIN
What made our material evaluation rewarding was the fact that hourly Production employees in two business units, with the assistance of a Quality manager and, later, the product engineer, drove it.

The team decided to run a test to compare the performance of the new material to the prior material. That was no easy task since they only needed a partial order of the prior material. This request went through Purchasing, Product Engineering, Production management, all the way to the CEO. At every level, it was pointed out that the new material was “identical to the old” and represented a big cost savings to the business. And at every debate, it was also pointed out that the defect and scrap rates were significantly higher, and the production rate was slower when running this new material. Scrap and waste were counted, and historical production rates were documented and compared.

After two months, a small order (one skid) of the old raw materials arrived. (The Purchasing manager, of course, took pains to explain to everyone how expensive this material was, and especially in such a small order.)

The test revealed a big difference in the two materials. A skid of the old (originally used) material ran with one feeding jam, fewer setup pieces of scrap, and no defects. The skid of new (less expensive) material ran quite differently: with numerous jams, extensive setup waste and multiple defects throughout the run. Yet, in the face of such results, Purchasing continued reminding the team that the new material was less expensive than the old, even factoring in scrap and waste factored. We responded by pointing out the matters of slower processing and delays in the production schedule.

Both skids of processed material were then moved to the downstream assembly operation for comparable test runs. Again, there was a day and night difference in how the two materials performed. The original material ran at design speed with fewer jams. The newer material required slower speeds and much operator time in tending the feeding section of the assembly process.

ROOTING OUT THE PROBLEMS
Further analysis revealed the new material was made mostly from recycled fiber, with almost no virgin content. The original material was 50% recycled and 50% virgin material. It had a grain structure that would maintain flatness. Conversely, the new “bargain” material had virtually no grain structure and would be warped in multiple dimensions, which caused setup problems, feeding jams, and defects.

Despite these findings, from a Purchasing perspective (inside the Purchasing “silo,” so to speak), it was still a better deal to use the new material to save money. However, looking outside that silo at the entire process (from start to finished goods) revealed the opposite. The cost of finished goods was higher with the less-expensive raw material. Moreover, meeting the production schedule was erratic.

As for the skid-count inaccuracy with the new material? The original supplier labeled its skids with the actual count. The new supplier labeled each of its skids with the total-order count divided evenly by the number of skids shipped. When those new- material skids were set side by side, the difference in height was obvious (in some cases, a 5- to 8-inch difference). Yet the tickets for the new material didn’t factor in any of those differences.

When the new material was processed, any setup waste and scrap was subtracted from the count on the skid ticket. Not a problem here. But when a skid of the processed materials arrived downstream for a specified number of products to be run, the inaccuracies would either stop production short or cause an unknown number of processed materials to be sent back to inventory holding as an uncounted short stack.

THE BOTTOM LINE
As RAM professionals, we should think beyond the PF-curve mentality and consider any and all types and causes for equipment unreliability. We are, after all, detectives of sorts, not only of maintenance-related equipment problems, but anything that prevents equipment from doing what it is intended to do, the first time every time. Sometimes, there are upstream and downstream issues that cascade into unreliability at multiple machines. Understanding and following the production process flow and related inputs at problematic equipment can reveal surprising causes that engage parts of the organization typically left outside of our reliability discussions.

Oh, one more thing (as Columbo would say): Regarding issues with the new material, some executives were known to have opined,  “That’s not a problem. We’ve built the cost of downtime, scrap, and waste into the standard cost of production.” In my own opinion, that, too was a big waste.TRR