by Andrew Sloley
At some time or another, nearly every engineer gets ensnarled in a difficult or complicated solution to a simple problem. The following story vividly illustrates this point.
The lead process engineer for a refinery revamp wanted to brainstorm about how to control the reactor inlet temperature. The conceptual design relied on a steam preheater to heat a hydrocarbon stream to the desired inlet temperature.
Headers for both 100 psig saturated steam and 250 psig superheated steam were available nearby. Unfortunately, the 100 psig steam couldn’t achieve the needed temperature. So, the superheated 250 psig steam at 520°F was selected. However, lack of attention to detail in conceptual design can hide many sins.
By using the superheated steam, the surface temperature of the tubes in the preheater would exceed the desired process temperature. This high temperature might lead to the formation of fouling products and reduce the downstream reactor run length. The next solution was to use condensate to desuperheat the 250 psig steam. The saturated 250 psig steam still would meet service requirements while eliminating all risk of high tube-metal temperatures. However, the closest available condensate was more than 1,000 ft. away. Getting it to the reactor would require running a dedicated line with freeze protection — at an estimated cost of more than $100,000.
The design is simple and elegant, but a but of reflection led to an even better solution.
The immediate question that came to mind was: “Why pipe condensate to the exchanger when we make condensate in the exchanger?” This quickly lead to an elegant design for an exchanger with an internal desuperheater (Figure). We had seen past the obvious but difficult solution (lots of line and a desuperheater) to a simple one (an internal desuperheater).
After a few minutes, another thought arose. Why did we have to desuperheat the steam at all? Perhaps we suffered from trying to make the problem too complicated?
As every engineer should know, desuperheated steam often isn’t. Heat loss often results in the superheated steam arriving at the process unit as saturated steam. Why not take advantage of this?
This gave us the idea for an even simpler solution: Take the heat conservation insulation off the saturated steam feed line. The remaining personnel-protection insulation allowed enough heat loss that the 250 psig steam would arrive saturated nearly all the time. For this application, that would suffice.
The moral: Even when you think you have found the best solution, put the problem aside and then ponder it again. A fresh look may lead to an even better solution.
As for the integral desuperheater, it is too elegant a solution to go to waste. Perhaps you may have a use for it.
Andrew Sloley, contributing editor [email protected]