While Process and Instrumentation Drawings (P&ID) often are neglected, Process Flow Diagrams (PFD) sometimes are treated as unnecessary and completely ignored. Many engineers don't appreciate their value.
PFD and their predecessor, Block Flow Diagrams (BFD), present a process in terms of equipment — without the clutter of control details given in P&ID. PFD and BFD often contain material and energy balance information, which could explain the reluctance in updating or even developing them: concern for technical security.
PFD and BFD, unlike P&ID, clearly depict a process flow. While cluttered drawings might not markedly impede understanding of a simple process, they certainly can obfuscate complex ones. You never would show a P&ID to an unfamiliar client or boss — it's too much to take in. Batch processes, especially ones involving a mix of batch and continuous steps, frequently are more difficult to explain on P&ID.
Another useful function of PFD is to present shutdown and startup details (often via dotted lines). They easily can become lost on P&ID.
Simplifying P&ID isn't the answer. Attempts I've seen in the pharmaceutical and refining industries, while successful, have sacrificed information that's sorely missed by control engineers. When these details are added drawings often are broken up so that processes remind me of a Picasso rendering of a clock — all the pieces are there but it doesn't make sense. Take a month off to try to follow such P&ID with fresh eyes!
PFD are most useful when they include a material balance. However, such balances often include proprietary information. So, here're some thoughts on developing a PFD that will be useful without giving away trade secrets.
First, consider ways to present data. Is it really necessary to include complete stream compositions? Why not only put basic information in the PFD and keep the detailed material and energy balance separate under lock and key? Or perhaps express the compositions as ratios. One textbook on catalyst beds reports PFD stream components in pounds of catalyst consumed per barrel of feed. For gases, why not show the stream molecular weight? For liquids, why not give the specific gravity or density, the viscosity or some other parameter? The information provided on the PFD should be useful without revealing too much. For example, a vendor can size a compressor by assuming ideal gas law if a molecular weight is supplied. (The gas volume usually will shrink significantly when non-ideal gas properties are assumed.) Likewise, someone can select an agitator by knowing the viscosity and density of the final mix. Having specific details about composition is unnecessary in many vendor packages.
Second, don't show details of process control on PFD. Why not simply indicate a meter and control valve for flow? There's no need to give away any secrets by identifying the type of valve and flow meter or that it's part of a cascade loop.
Some data shouldn't appear in PFD. Don't include information acquired empirically or that would give competitors insight into the process. When I worked as a researcher in the Air Force the most guarded detail wasn't the rocket propellant formulation but the steps required to manufacture it. It would have been amusing to watch the Russians try to duplicate our propellant from the formulation. Sometimes a competitor won't gain that much from the process steps. For instance in a fermentation, details like what nutrients are added and when are crucial.
Keeping such information off the PFD doesn't mean adding it to P&ID — it would worsen the clutter that already exists on most P&ID.
The complicated nature of batch processes or those that include batch steps favors using a PFD over P&ID. In most cases the same instruments and controls will serve a number of batch products. The best way to represent these processes, I believe, is with a PFD for each batch process with a separate P&ID only if warranted by a different control scheme. Exercise special care with batch processes because of their proprietary nature.
Dirk Willard is a Chemical Processing Contributing Editor. You can e-mail him at firstname.lastname@example.org.