He won a bonus for his suggestion. Diluting the yeast prior to centrifuging increased the product concentration. But the yeast cells absorbed more water, which required higher salt concentrations in the packaging department; salt "sweats" water out of the yeast. Higher salt usage meant faster accumulation of toxins, like mercury and cadmium, in the salt brine tank. This waste was removed by vacuum truck. In addition, the salt contributed to soil damage at the farm where wastewater was used to grow crops. Obviously, this "improvement" hurt the plant's bottom-line. Too often, a suggestion is carried out without proper analysis of its implications.
One young engineer was told to replace the carbon steel bolts on the cover of a clamp-filter. He must have thought, "Oh, just use stainless steel bolts." By the time the project was handed to me, a box of type-316 stainless steel bolts, nuts and washers was waiting. I was told, by the same young engineer, now manager, to close the book on the project. Oh, if life were that simple. It never hurts to spend some time before starting a project by thinking about it. If possible, get together with someone and consider how the project impacts, and is impacted by, other activities in the plant.
After a few moments thought, I saw a threat my young boss never imagined. To improve my chances for acceptance, I called the mechanical integrity group. I asked a mechanical engineer, What do you think about replacing carbon steel fasteners with 316 stainless?" I already knew the answer, of course, but I wanted him to be on the same page. "We'll have to de-rate the vessel (filter) design pressure by about 15%, if I had to guess," was his reply. "Oh, then, I should ask production if they can live with (doing the numbers in my head) a 40-psig rating?" I quickly called production and received a, "Hell no!" After some design work, by committee, we came up with an acceptable design -- a hard 400'series steel for the bolts and washers and a soft 400'series steel for the nuts. This solved the problem, although I think the young manager was embarrassed.
How can you avoid this kind of embarrassment? First, define the problem. If the definition is not something you can measure or rationalize, keep asking questions -- quantify. For example, "Why increase the concentration of the yeast?" Which leads to "How do I increase the concentration?" Next, list the potential solutions, with pros and cons. Use this table to predict relationships between solutions and problems. A simple prolcon table almost certainly would have revealed the drawbacks to adding more water and would have pointed the production manager in a different direction. Usually, several solutions will be provided -- cheap solutions should be implemented regardless while expensive ones require selection.
Economic analysis is not often useful at this point, because environmental and safety aspects are difficult to quantify. Once a problem is defined, and solutions have been selected, create a flow chart. Using the flow chart, develop an interrelationship diagram. This will show you how items in the flow chart relate to each other. When the interrelationship diagram is complete: revisit the flow chart. After preparing a final flow chart, have someone outside the group, if there is a group, look at it to provide a fresh perspective to avoid "group-think." This phenomenon occurs when people who work together begin to think the same. A sort of suppression of criticism sets in, which is ineffective, and where safety is involved, is dangerous.
Don't get locked into a methodology, be creative! When the problem is one of substitution, like the replacement of carbon steel fasteners, try a comparison chart -- list all the properties of the item and compare them against the replacement.
And, here is the most important thing to remember: check your hypothesis later! Problems often develop a life of their own. Course corrections will become necessary. Not only can the flow chart be expected to change, to become more complex, the interrelationship diagram will also grow more complicated. This is normal. The more you know about a problem, the more complicated solutions often become.