July 2020


Reliability: Finding a maintenance cost optimum

An Italian company is among those that design and build industrial machinery for the tissue industry.

Bloch, Heinz P., Hydrocarbon Processing Staff

An Italian company is among those that design and build industrial machinery for the tissue industry. One of the company’s service engineers was reading a book on machinery uptime1 when he asked us to clarify the concept its authors had tried to convey in FIG. 1. What puzzled the reader was an accompanying statement to the effect that maintenance and operating costs decrease as initial capital increases, up to a benchmark point of maintenance costs.

FIG. 1. Minimizing total lifecycle cost, top curve.

We formulated a reply, using a simple, two-point analogy with airplanes:

  1. Suppose an Airbus 380 costs 10 times what an inter-city/regional plane costs. However, the Airbus 380’s “per-passenger mile flown” cost of maintenance and fuel is much lower than what the airline pays for the smaller plane’s maintenance and fuel.
  2. Smaller machines are generally less efficient. Per kilowatt-hour (kWh) of power consumed by smaller machines, its owner-operator probably pays more for maintenance and power; therefore, the curve is high on the left side and much lower on the right side.

However, FIG. 1 is only an arbitrary example curve. In the case graphically depicted here, it would be a machine that costs $40,000 if it has small throughput capacity and $150,000 if the larger model is purchased for high throughput capacity. In this simple illustrative example, we attempted to demonstrate the well-known principle of “economy of scale.” However, this graph can also demonstrate “economy of value,” as will be explained.

What the low-cost mixer gearbox really cost

An actual field example involving a mixer gearbox is fully described on a spreadsheet in the referenced text1; the example dates back to 1992. When gear-tooth pitting after the first year of operation prompted a reliability engineer to conduct a failure analysis and design review, it was evident that the American Gear Manufacturers Association (AGMA) service factor was only 1.07. However, American Petroleum Institute (API) standards require a 1.70 service factor for equipment associated with extrusion machinery.

The investigating reliability engineer specified a new gearbox with a 2.0 AGMA service factor. The gears were wider and the gearbox bigger; the motor had to be moved several inches. The cost of the new gearbox and motor relocation was $294,333; the original gearbox cost $160,000. The new gearbox ran for 5 yr before requiring significant repairs. Cost performance was compared on a 10-yr basis, with future years assumed to require the average maintenance costs of the known years.

The spreadsheet1 showed the lifecycle cost (net present cost) of the 1.07 service factor case as $484,023 vs. the lifecycle cost (net present cost) of the 2.0 service factor case of $345,756. Obviously, the 2.0 service factor case wins. The stronger gearbox should have been selected at the beginning. The difference would have been even greater if production losses had been included. However, the example as stated illustrates how lower lifecycle costs may be achieved despite higher initial cost.

Avoid cheap equipment

In summary, it can be easily seen that the lowest cost of initial capital may not produce the lowest lifecycle cost in the long run. Maintenance and operating costs can be quite significant and cannot be ignored in an economic analysis.

Looking at only the initial cost of a machine, or emphasizing only its better efficiency compared to competing offers, is usually a serious mistake. Conducting a thorough analysis and allowing a reliability professional to take ownership of a selection process is valuable. These initiatives often ensure the long-term continuity and staying power of a commercial entity. HP


  1. Bloch, H. P. and F. Geitner, Maximizing Machinery Uptime, Gulf Professional Publishing, Houston, Texas, February 2006.

The Author

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}