This Month’s Puzzler
We use liquid hydrogen cyanide (HCN) in our process. It’s stored in a 150-psi-rated tank maintained by nitrogen at 45 psig. Although we have relief valves with rupture discs, we rely on a 1-in. pressure-reducing ball valve (reduced port) for most situations such as railcar loading; the Cv = 28. The reducing valve opens at 55 psig, with a discharge pressure of 1.5 psig. Our vent line is 248-ft long. It’s 2-in.-dia., schedule-80, 316L-stainless-steel pipe and has butt-welded joints; flanges are ANSI rated for 300 psi. Barometric pressure is 14.35 psia. The operating temperature is 86°F. Vented gas from this valve goes to a 4%-NaOH venturi/scrubber and then to atmosphere; any flow from the relief valve also goes to the scrubber. We lost the original calculations for the vent valve, which was obsolete. That valve was 2-in., with a Cv of 115. We estimated the gas flow at 5,300 pph (80% N2 and 20% HCN), based on hand calculations. Now, the relief valve pops too often. In addition, sometimes our scrubber seems to be overwhelmed, and we must write an emission report for HCN. Did we size the new valve correctly? What’s up with our scrubber?
Change The Valve Immediately!
A back-of-the-envelope calculation for the Cv shows that the 1-in. reduced-port ball valve will work for noncritical flow. However, the equation for choked flow indicates that a Cv larger than 28 is required when that occurs:
Cv = Qgsc×(SGsc×T)½/816×P1 (choked flow)
Cv = (Qgsc/962)×(SGsc×T/(P12 –P22))½
where Qgsc is standard cubic feet of gas flowing at 60°F and 14.7 psia; SGsc is molecular weight of gas/molecular weight of air; T is in °R; and P1 and P2 are in psia.
I calculated a Cv of 26 for noncritical flow and a Cv of 29.9 for choked flow. The Cv of the 1-in. valve, while adequate for 5,300 pph if noncritical flow existed, is slightly inadequate for choked flow. Choked flow actually occurs because the valve discharge pressure, P2, is less than 51% of P1.
It is important to note that these back-of-the-envelope calculations assume an ideal, dry gas. However, because the HCN vapor coming off the storage tank will be in equilibrium with the liquid, the stream will be close to saturation — so it’s definitely non-ideal vapor. I usually prefer to use ideal gas calculations because they tend to be 1–5% more conservative; I like to hedge my bets.
Dropping the pressure through a control valve will lead to a slight Joule-Thompson temperature decrease, causing condensation. Even 10–15% entrained liquid after the valve will increase the pressure drop in the vent, resulting in a higher valve discharge pressure and a greater required Cv. It looks like a 1-in. valve was a poor choice.
As for the emission reports, it seems the calculated relief flow is too low. Relief valves are required to be slightly over-sized. So, you also should review the sizing criteria for your relief valve.
The vapor velocity for a scrubber is one of the criteria for sizing. If the flow is too great, the volumetric mass transfer coefficient (KGa) for the packed bed will drop. Because you can’t add more packing, consider raising the caustic irrigation rate as high as possible.
Maybe you can increase the impeller size for the recirculation pump without exceeding the motor limit or you could replace the pump altogether; generally, motors are cheaper than impellers — sometimes, it’s best to swap the whole pump with a larger one, though. Just be careful about the pump base and piping.
Dirk Willard, consultant
We just started toll manufacturing a synthetic lubricant but had to end production abruptly. As we do for lots of chemicals we produce, we use process equipment that previously had made other products. In this case, we employed a vacuum tower filled with structured packing that recently handled distillation of omega-3 fish oil. It was cleaned with steam and detergent before our campaign.
When we started up, we saw a poor cut in our reflux tank before the vacuum pump — so we maxed out the reflux ratio and increased the reboiler temperature. Shortly after that, we killed flow to the tower because the vacuum pump tripped on over-current. Then, after the shutdown, we noticed a cracked elbow in the PVC chilled water line feeding the condenser. When inspecting the tower, I discovered a toluene vapor cloud. Our plant manager is convinced the vacuum pump failed because of poor maintenance and dealing with that issue will end our problems. His other theory is that the plate-and-frame condenser may be leaking because of gasket contraction. I’m wondering if he got this right. What do you think?
Send us your comments, suggestions or solutions for this question by November 11, 2016. We’ll include as many of them as possible in the December 2016 issue and all on ChemicalProcessing.com. Send visuals — a sketch is fine. E-mail us at ProcessPuzzler@putman.net or mail to Process Puzzler, Chemical Processing, 1501 E. Woodfield Rd., Suite 400N, Schaumburg, IL 60173. Fax: (630) 467-1120. Please include your name, title, location and company affiliation in the response.
And, of course, if you have a process problem you’d like to pose to our readers, send it along and we'll be pleased to consider it for publication.