Design & Simulation

Plant Design: Reduce Revamp Risks

A successful project requires attention to several crucial preliminaries

By Andrew Sloley, Contributing Editor

Some sites have so many problems with revamps that the very word “project” can cause plant managers to dig in their heels to defend the status quo. “If it’s not broken, don’t fix it,” or something similar, they exclaim. Unfortunately, resistance to change can endanger long-term viability by preventing plant improvement. However, critics of change have a rational basis for concern. Many projects do fail. In fact, some studies show that more than half of plant revamp projects don’t meet some combination of schedule, cost and return-on-investment criteria.

A cost estimate target sets the level of work possible, and vice versa.

Successful change isn’t accidental. It comes from consistent application of basic principles and understanding the plant and its objectives. So, let’s examine some parts of successful change.

At some point the project must move from a general idea of let’s do something, to particular plant modifications. The current emphasis in project management is to use a staged-gate approach. Multiple steps add more detail to the project and define current estimates of costs and benefits.

Want to improve the design of procedures? Register for this free Process Safety  webinar now!

The most common source of project failure is unwillingness to accept, or even understand, the costs of preparing realistic cost estimates. These costs include both the duration of the effort (schedule time) and the scope of the effort (engineering time). This problem is nearly completely the responsibility of the plant owner and its management team. If you work in a plant where every project “fails,” the odds are that decisions by your own management lie behind the problem. Your cost analysis is only as good as the data behind it. If you aren’t prepared to gather and generate good data, you should expect to fail. (Other issues also can undermine success, see: “Win by Planning to Fail.”)

Beginning with the moment where a concept moves into putting together a good design package, what will your team need as a starting point?

First, the team needs a project definition. What are the project objectives? What is the plant prepared to pay to achieve these objectives? What tradeoffs between objectives are acceptable? Also, be specific about the cost of doing nothing — this might be the best solution for some problems but for others might mean going out of business.

Second, define the current scope of work — either via a specific cost target or the level of work the plant is willing to pay for. Don’t define both. If you opt for a cost estimate you have set the work needed, and vice versa. Unwillingness to accept this point likely will guarantee your project will fail to meet objectives.

Third, gather plant information. Necessary background information depends upon the scope of work defined. For almost all significant projects, this includes plant heat and material balances linked to process flows as shown on process flow diagrams. These balances should detail both process and utility loads. They also should balance within realistic limits of accuracy for plant instrumentation. You will need piping and instrumentation diagrams and equipment information for anything more detailed than a cursory analysis. Also, include stream analytical results from the operating unit.

If you consider your process confidential, it’s certainly okay not to disclose all available information to an outsider. However, not sharing such information means you assume any risks and uncertainty that this creates.

Depending upon the level of detail required, you might need additional information such as plot plans, location of underground lines and supporting infrastructure. Older plants often lack up-to-date details on underground lines and utility infrastructure. Updating this information may be essential. Don’t underestimate the potential cost implications of ignorance in these areas.

For specialized processes, thermodynamic and physical property data often are critical. If your plant is the only one making a specific product, you should plan to provide these data. Defaults provided in software tools simply aren’t good enough for many engineering evaluations. Software gives great precision; accuracy may be guesswork.

Adequately attending to these steps, plus accepting that accuracy in cost estimates costs money, will provide a sound basis for making rational project decisions. This is the key step in having a successful project.

Sloley2ANDREW SLOLEY is a Chemical Processing Contributing Editor. He recently won recognition for his Plant InSites column from the ASBPE. Chemical Processing is proud to have him on board. You can email him at