1
Focus of Book

1.1 Introduction

The book is focused on what technical concepts and calculation-based approaches are required to improve and maintain Operational Excellence. The types of problems that are envisioned are either chronic operating problems or operating problems for which the problem solver has no experience-based solution. The example problems discussed are for the most part real-life problems that actually occurred, and solutions were those that were developed from a careful analysis and calculations. These problems and solution concepts are enhanced by a limited discussion on setting metrics and uncovering hidden problems.

While if experience-based solutions are available, they can be of great value. However, not every problem encountered has been experienced before by the problem solver, and as such these problems are rarely solved with experience-based solutions. In fact, the attempt to solve them based on extrapolated experience often results in delaying the true solution. For example, an impurity problem associated with ethylene used in a polyethylene process cannot be extrapolated to a similar impurity problem associated with propylene used in a polypropylene process. While this seems obvious, time pressures associated with poor operating performance often create the management decree to “Do something even if it is wrong.”

In reading Sir Arthur Conan Doyle’s late 1880s fiction about the famous detective Sherlock Holmes and his companion Dr. Watson, I see a parallel to current industrial problem solving techniques. Watson often quickly jumped to conclusions based on a few facts and his experience. On the other hand, Holmes took a thoughtful and methodical approach and used all the available data to obtain the right solution. Holmes’ approach being methodical appeared to be too slow, but it avoided the endless detours caused by jumping to conclusions. A quote from Sherlock in the novel The Valley of Fear illustrates this well—“The temptation to form premature theories upon insufficient data is the bane of our profession.” For the complicated or chronic process problem, the successful problem-solver is a modern-day Sherlock Holmes.

1.2 Metrics and KPIs

Before one decides that a problem exists, there must be a deviation from some sort of standard or metric. These metrics are often referred to as key performance indicators (KPIs). This terminology KPIs will be used in the book to refer to the anticipated value of a key variable. An Operational Excellence deviation is one that is significantly different than the anticipated value of a key operating variable or lumped parameter. The key variable or lumped parameter can be as simple as pressure of a key utility or as complicated as a reaction rate constant. Before one determines that a key operating parameter has significantly deviated from the target value, the target value must be known. The reported analysis of blood of a medical patient uses this concept well when the technician reports actual values and the expected range.

The first step in problem-solving is to determine values for the KPIs. The approach of employing a lumped parameter should be utilized. For example, the lumped parameter for a heat exchanger is the well-known heat transfer coefficient. This includes temperatures of both hot and cold sides, flow rates of both hot and cold materials, and thermal characteristics of the materials in a single value. A similar approach could be used for a reaction rate constant. The calculation of this lumped parameter will provide a value of this KPI at any point in time, which can be compared to the theoretical value. In the case of a heat transfer coefficient, the theoretical value is the design coefficient. This comparison can then be expressed as a percent of the theoretical value. A similar approach can be used for equipment such as pumps, where the pump curve serves as the theoretical data source. For example, a key pump operating at the design flow rate, but below the theoretical head, could be flagged as an operating problem using the design head as the theoretical KPI.

Regardless of the terminology (metrics or KPIs), the most important thing is how the theoretical value is determined. It can be determined by design calculations such as heat/material balances, equipment specifications, or pilot plant/laboratory results. They can also be set by actual plant performance. It is often true that process plants can be operated at 10% or more above design capacity for extended periods of time. This capability should be used to determine new KPIs rather than using design calculations.

These KPIs must be theoretically obtainable, achievable (with often significant effort), and sustainable. There is often a temptation to set the KPIs too low so that they can be easily obtained. This temptation must be avoided if the goal is to achieve Operational Excellence. The successful military mission or airplane journey represent examples that are theoretically possible, achievable, and sustainable.

Now considering this terminology in more detail:

• Theoretically obtainable KPIs are those set by the physical limits of the operating unit. These are metrics that would be achieved if there were no limits caused by failures, errors, or uncontrollable events.
• Achievable KPIs are those that can be achieved with existing equipment and available raw materials.
• Sustainable KPIs are those that can be achieved every day of the year. Again, referring to the airline industry and assuming that one is traveling from Tokyo to DFW (Dalllas-FT. Worth International Airport) with an exceptionally strong tail wind, the flight may be as much as an hour shorter than planned. However, this does not mean this shorter flight time is sustainable day after day.

KPIs can include things such as operating rates, reaction kinetics/rate constants, heat transfer in critical exchangers, tower performance, and/or dryer performance. To achieve improved Operational Excellence, these KPIs should be established during good plant operations. The approach to establishing them during good operation frequently at greater than design capacity will often create concerns about the number of chronic problems that are uncovered. This uncovering of what are referred to as hidden problems should be considered a valuable source of potential operational improvements. The other extreme of setting metrics so low that they are easily achieved is not the focus of improved Operational Excellence. While setting KPIs too low does reduce the number of perceived operational problems, it does limit the KPIs or however metrics are measured to levels that can be easily achieved. A quote from Frederick Bonfils, the founder of The Denver Post, is certainly appropriate for the manufacturing industry—“There is no hope for the satisfied man.” It is only by setting high standards and tight KPIs that Operational Excellence can be improved.

1.3 Finding Hidden Problems

Comparing these theoretical KPIs to actual operation is a method for finding “hidden problems”. Finding these problems is one of the keys to improving Operational Excellence. These hidden problems exist for a multitude of reasons. They can be understood as operating problems that have become accepted and hence ignored as “just the way that it is.” Since these problems are accepted, the necessary analysis to fully understand and to eliminate the problem is not done. Some typical areas are as follows:

• It is often difficult to convince a process control engineer to make a programming change in a process that is computer-controlled. This results in less than excellent adaptive solutions.
• In processes that are not computer-controlled, operators at shift change often “tweak” set points to their personal preferences. While the process will operate at either set point, one of the set points will be less than optimum.
• Often a process engineer does not understand mechanical equipment such as pumps and compressors. This lack of understanding limits the ability to make any potential recommended changes that might improve process operations. The process will operate, but not at an optimum.
• There is confusion between using standard deviations for control and using them for defining if a situation is likely a real problem. The well-known ±3σ has become a benchmark for process control. However, for operating problems, problem-solving should begin before that point. This concept along with the risks is described in a later chapter.

These hidden problems or ones that are and have been blatantly in the open for either short or extended periods of time are referred to as chronic problems, and solving them is what the book covers. These operating problems are almost always related to some aspect of technology. This technology could be equipment technology or process technology. Unfortunately, attempts to solve these problems are often based on operating experience rather than an understanding and analysis of the problem. They are defined as chronic because these attempts have failed to resolve the problem. Many of the example problems included in this book fit the definition of chronic operating problems. They were long-standing problems with many trial-and-error attempted solutions. The attempted solutions were based on limited experiences, intuition, or guesses. The time period to attempt these multiple solutions almost always exceeds the time that would have been required for a more in-depth analysis of the problem...