Avoiding Shutdowns with Accurate Level Measurement

For plants and mills that use steam in the production process, measuring the steam condensate level accurately can be the difference between operating reliably and fighting unplanned shutdowns. So was the case for a U.S. ethanol producer, Absolute Energy, as told in a story by ConnectingIndustry.com’s Automation magazine.

The story, Technology resolves costly shutdowns at ethanol plant, describes an application with a condensate receiver tank on a steam heat exchanger. For those that don’t use steam in their production process, these tanks collect water that has condensed from steam and return it for reuse, divert it to storage, or discharge it for wastewater treatment.

In this application, level control was critical:

…to continuous operation of the heat exchanger and molecular sieve dehydration process. If water levels get too high in the receiver, temperatures in the heat exchanger drop, which can cause wet ethanol to enter the molecular sieve beds forcing the process to shut down and the ethanol-water mixture to be reprocessed.

For this plant, shutdowns cost €13,500 ($19,500 USD) per hour. After a shutdown, which occurred every few weeks, it takes 20-30 minutes to restart the process. The startup consumes additional energy over and above steady-state operation. If the sieve beds get wet:

…the process shutdown lasts between 36 and 48 hours because the beds must be shut down, regenerated, and restarted. While this type of upset is not common, wet sieve bed occurrences eventually lead to extensive equipment damage that combined cost over €135,000 [$195,000 USD].

The existing level measurement was based on differential pressure (DP) measurement. Pressure changes caused wide variations in the “wet leg”. The changing density of the condensate also impacted the measurement. Together, these pressure and density changes led to level measurement inaccuracies, which caused the level in the tank to rise too high. This in turn, caused:

…wet ethanol to enter the molecular sieve beds forcing the process to shut down and the ethanol-water mixture to be reprocessed.

The solution was to replace the DP level measurement with guided wave radar (GWR) technology. The accuracy of the guided wave radar technology is not impacted by changing pressure and density conditions. The project team installed a Rosemount 5301 Guided Wave Radar and 9901 Chamber. The chamber provided a way to mount the GWR externally to the tank.

The article sums up the project results:

The improved control of levels in the condensate receiver means this measurement point is no longer a maintenance headache. As a result, Absolute Energy has increased plant availability, decreased energy and utilities costs, reduced maintenance costs, and lowered risk of equipment damage.

MP3 | iTunes

Audio clip: Adobe Flash Player (version 9 or above) is required to play this audio clip. Download the latest version here. You also need to have JavaScript enabled in your browser.

Update: I fixed an inaccuracy in the wet leg sentence which incorrectly implied a remote seal system.

2 comments

  1. we are having an issue with one application. a steam and condensate vessel that receives a condensate back from a steam heater (heat exchanger). the issue is, was noticed by operations team that when vessel level (dp transmitter) goes above 90%, hammering occurs due to steam condensate mixup. my question is around the proposed solution which is to rescale the range of the dp level transmitter, lower the high elevation tapping and recalculate the range (all of this will be in system, configuration only) . will this solve the problem ..please advise.

  2. Jim Cahill says:

    Hi Yassin, Thanks for your comment and question. I checked with my friends who manage the Rosemount DP level measurement line of products. They explained that since the hammering only occurs when the level is above 90%, lowering the top tap would lower the range points and therefore lower the span–which will prevent the hammering.

    Another easier method would be to just lower the span or 20ma point of the current transmitter–therefore the controlling ranged would be lower.

Leave a Reply