Improving return on industrial assets
If managers cannot measure performance, they cannot effectively diagnose and resolve problems. Overall equipment effectiveness (OEE) has become a critical key performance indicator (KPI) for companies seeking to measure asset performance and improve operational decision making.
In essence, OEE is a compound KPI derived from three to four factors that measure the equipment’s operational performance—that is, how well a manufacturer converts its time into output. OEE can play an important role in the drive for industrial high performance, but to be effective it needs to be approached systematically and holistically, and involve people, technology, and processes.
OEE is a product of three factors: A, P, and Q. Availability (A) looks at whether the machine is running or is stopped because of failure, maintenance or other scheduled breaks. Performance (P) evaluates the equipment’s actual production rate in comparison to the rate it was designed to deliver. Quality (Q) assesses how well the items being output by the equipment are meeting specifications. With these three factors considered, OEE can be a useful tool for discovering a plant’s hidden potential.
Each factor covers one dimension of equipment operation; thus, OEE = A × P × Q. Although overall equipment effectiveness is relatively simple to calculate, in Accenture’s experience it is not necessarily a universal plug-an-play KPI. A number of products on the market provide standard OEE formulas, but OEE needs to be approached from a holistic perspective, and tailored to the individual situation.
Given the numerous variables within a company, a broad yet comprehensive methodology helps companies apply OEE management to any industrial process. Accenture’s methodology for OEE, as shown in the Methodology Model, defines the proper OEE indicator and underlying factors for each specific case.
Applying OEE analysis
The question then is: What kind of benefits could a company expect from such a methodology? To find out, it is useful to look at a specific example: the application of the methodology to a continuous casting machine in a plant operated by a major steel company. An OEE management software module was designed, configured, and used to calculate the overall indicator. This enabled an analysis, a recommendation of changes focused on the identified major cause of loss, and an estimate of the possible increase in the company’s return on assets (ROA). ROA is the ratio between the organization’s operating profit and its total asset average.
Shown in the Methodology Model, Step 1 designs the OEE tool’s data model. The most important points to consider here are:
• Flexibility—the indicator and its factors’ names and formulae must be customizable.
• Equipment status modeling—productive time and downtime must be clearly defined so that there is no overlap.
• Standardization—the data model should be ISA (International Society of Automation) 95 compliant so that it is in line with market trends and easily applicable to different industrial processes.
When it comes to acquiring process data (Step 2), automated solutions and systems integration play an important role because they reduce inaccuracies and human error, such as typing mistakes. Technologies typically used in this area include OPC (Ole for Process Control, used for open connectivity), Web services, and message queuing.
The bar chart shows a graph generated during an OEE indicator and factor analysis for the continuous casting machine applied over a one-month period. Each factor was calculated as:
Availability = total productive time/total time.
Performance = average real speed/theoretical speed.
Quality = total good production/total production.
Comparison and drill down
Analyzing and comparing the results to a typical world-class OEE calculation, it was clear that the operation was running below ideal KPI levels: Availability should be greater than 90 percent; performance, greater than 95 percent; and quality greater than 99 percent. Since two of the three factors were low, a drill-down analysis was conducted to identify hidden potentials.
When talking about a machine’s performance, a drill-down analysis typically compares theoretically estimated and real production amounts and speeds. In this case, the estimated production amount for the month was 81,400t of steel, while the actual amount was 54,900t. Nominal speed (as obtained from the company’s master data) was 133.33t/h, while average real speed measured reached just 89.945t/h. If such low performance rates persist for a couple of months, one can conclude that the master data may not be reliable or may not be updated frequently enough, pointing to a need for process definition studies. The question here is how fast are the machines really able to operate?
The results for the quality factor, on the other hand, indicate a different issue. During the month in question, 54,500 tons of steel met quality standards, while 17.06 tons needed rework (or, as the company literature says, was not "first time right") and 382.28 tons became scrap. This means that almost 100 percent of production was meeting quality standard, which experience shows to be unlikely. In the case example here the explanation for the quality factor patterns was that the products differ from each other only in quality requirements. Thus, making product B when A was planned can be considered alright, since it also has high market value.
In Step 5, the focus is on identifying major problems and hidden potentials, using root-cause analysis and planning-improvement efforts. As mentioned above, the ROA indicator is commonly used for this kind of analysis.
Regarding the continuous casting machine in our example, performance was the major factor responsible for losses in operational effectiveness. However, it is not easy to define whether the problem is in the equipment, in the process, in the master data, or in a combination of all three.
So, a more feasible and focused improvement plan was proposed: Increase equipment availability by eliminating the seasonal break (7.71 percent of equipment time). In simulation, reducing the seasonal break and assuming that other stop causes keep the same pace (7.42 percent of this time span refers to equipment’s stopped time), the result is a real availability gain of 7.14 percent.
Thus, the new values for this context will be: availability: 92.01 percent (increased 7.14 percent); performance: 64.46 percent; quality: 99.27 percent; and OEE: 61.62 percent (increased 8.4 percent).
Effects of OEE on ROA
Increased uptime will obviously increase the production rate, but it will also incur higher costs. The table shows a comparison of hypothetical financial data for a production line before and after OEE improvement. Considering an original 10 percent ROA and the $9 operating profit, the total asset average is equal to $90. In this case study, a fixed asset amount over total asset amount of 70 percent was assumed. Therefore, the total asset coverage calculation (again using hypothetical data) is: Total asset average example = 0.7 × $90 + 0.3 × $90 × 1.084 = $92.268. The new ROA value then is $12.11÷$92.268 = 13.12 percent.
This means that an 8.4 percent overall equipment effectiveness increase would lead to a 31.2 percent ROA increase, at the cost only of additional workforce hours. Obviously, many other factors affect the return over investment. It may not be so easy to eliminate seasonal breaks, for instance, because of labor laws or prohibitive energy costs.
Based on our industrial experience, it is clear that aligning manufacturing management with business drivers is very important to an OEE effort. Defining relevant indicators that bring bottlenecks and root causes of loss to light is crucial, so efforts can be properly aimed at increasing production effectiveness in the current state—that is, without large additional investments in factory capacity.
To increase the effectiveness of equipment, it is critical to focus on creating the right OEE definition and measurements, and on having operators and managers perform rapid corrective actions. When approached correctly, OEE can play a vital role by providing the organization with the tools needed to improve plant performance and help an industrial company on its journey to industrial high performance.
Financial data comparison, using hypothetical data
|Source: Accenture Automation & Industrial Solutions|
|Carlos A. Laurentys is a manager with Accenture Automation and Industrial Solutions. He works with clients in the mining and steel industries on the implementation for best practises on production, quality, maintenance,and inventory management. Carlos is located in Belo Horizonte, Brazil. email@example.com|