Q & A > Question Details
There are 3 hydrogen plants in our refinery. All of them are steam reforming process. We use OG, LPG and naphtha as feed to produce high purity hydrogen. The raw hydrogen stream comes from steam reformer, after cooling down then sent to PSA to recover high purity hydrogen.
The design recovery efficiency of PSA is 89%, we found the actual efficiency at high throughput condition is 85% or less only. This is a bottleneck for hydrogen plants.
There are some comments from outsources. Someone said the operation life of molecular sieve used in PSA is very long, 10 years or more, we don’t need to replace it. But another had opposite comments. Would you please advise.
 
Answers
13/08/2012 A: Frank Lomax, CB&I, fdlomax@cbi.com
Adsorbent deactivation itself isn't the only possible cause for failing to meet your design recovery. Factors such as excessive operating temperature, migration of the adsorbent layers, channeling due to flow upsets during a bed switchover event; all could be the root cause. Perhaps the worst culprit in warm climates is high ambient or operating temperature, especially in steam reformer service, where the risks of poisoning due to unplanned for contaminants is not very high. It is possible that more modern adsorbents can deliver greater performance if an adsorbent replacement is desired, allowing the design performance to be met or even potentially exceeded. We are happy to review your operational data to assist you understanding the root cause of your problem.
03/11/2010 A: J Sudeep, SATORP REFINERY, jsudeep81@gmail.com
As you are aware, PSA is basically made of 2 to 3 adsorbents.
Efficiency may go down if the plant is operated in higher throughput, higher feed temperature, poison to adsorbents & operation factor. Check if any abnormalities during earlier stat up/shutdown activities and try to find the cause this can be done by collecting proper data.
I want to share one unfamilar experience which caused to drop in efficiency.
The feed was contaminated by heavier Hydrocarbon, as a result some waxy substances started to deposit downstream of shiftconvertor followed by PSA interconnecting pipeline. This caused reduction in plant throughput.
03/08/2010 A: Lindsay McRae, Pall Corporation, Lindsay_McRae@pall.com
Mol sieve / PSA life depends on how hard you push it, plus also contaminants in the feed gas can reduce mol sieve. Typical mol sieve / PSA life is up to 7 years in my experience when mol sieve is protected by a high efficiency Liquid Gas coalescer. If mol sieve is subjected to free liquid carryover, due to either entrainment or condensation (caused during cooling) then aerosol sized droplet quantity can be significant and this can lead to more frequent mol sieve regeneration, and short mol sieve life. Life can be shotened to <3 years or even down to 1 year if subjected to free liquid in the feed gas.
KO pots, Mesh pads, cyclonic separators, or other bulk separation devices like horizontal filter-separators are not able to effectively remove aerosol sized droplets. High efficiency LG coalescers such as SepraSol LG coalescers can achieve <0.003 ppmw free liquid downstream of the LG coalescer.
So if you are having short mol sieve / PSA life, it is worthwhile to check the performance of your inlet device, and if not of high efficiency type and liquid carryover is occurring, then add a new high performance LG coalescer to remove virtually all free liquid from inlet feed gas.
19/07/2010 A: Balasubramanian Iyer, Nexen CNOOC Limited, balasubramanian.n@aol.in
Following are the Reasons for Low Recovery in PSA :
1) If the Throughput of the Plant is high and the Kf is kept low then the recovery will also reduce.
2) See in the Past that the Methane Slip and CO Slip from Reformer and Shift Reactor Section was high for a long time and Kf was also not reduced during that period. If so these impurities would have plugged the Absorbents leading to Low Recovery. Try to reduce the Methane and CO Slip from the Shift Reaction Section.
3) If you are using other refinery gases also in your PSA section like CLPS(Cold Low Pressure Separator Off Gas from HCU, DHDT), CCRU(Continuous Catalytic Reforming Unit) check for their Composition mainly for higher hydro carbon, H2S and NH3. these can be a potential poison for the absorbents. If the composition of these are on higher side then reduce taking these gases in your PSA to increase Recovery.
4) Check for the Condensate carry over in the PSA inlet header. If Condensate carry over is high and not been drained properly this may also lead to the Low recovery of PSA. Check for the Demister pad dislocation/damage in the Upstream Process condensate Vessel in your next Shutdown.
5) Check the Off Gas of PSA and for its composition and product Hydrogen sample to check whether the Methane, CO and CO2 in the feed, off gas and in Product hydrogen is high or as per the design allowed value.
If the Above points are all checked then one can conclude that the adsorbent is affected and hence go for regeneration of the adsorbents or change over of the adsorbents.
The life of the Adsrobents are almost equivalent to the Plant Life ie 20 Years. As they are continuously getting regenerated by the process of De Pressurization.
Generally the PSA adsorbents can be regenerated if they are plugged with the heavy hydrocarbons etc by taking hot N2 and purging the gas to flare for some time by checking the composition of the Purging N2 Gas.
13/07/2010 A: Balasubramanian Iyer, Nexen CNOOC Limited, balasubramanian.n@aol.in
There may be many reasons for such low recovery of PSA and we cannot directly go into the activity of the absorbents.
1. Is the CH4 slip at the steam methane reformer is high than the actual as per the design feed to PSA? Compare the CH4 slip with the corresponding Steam-Carbon Ratio and the outlet temperature one can conclude whether the CH4 slip is high or not.
2. Is the CO slip from the Shift Section is high above the design feed condition to PSA? This can be verified from the laboratory results of the sample collected at the outlet of the Shift Section Reactor.
3. Is the Kf value of the PSA kept lower to ensure product hydrogen purity? If so at lower Kf the recovery will come lower due to the Decrease in Sub Cycle time of the PSA and hence loosing Hydrogen in the Off Gas.
4. If you are using any other Refinery Gases like CLPS (Cold Low Pressure Separator Off Gas) from Hydro Cracker, Diesel Hydro Treater, or Catalyst Reforming Unit gases in the PSA then check for its Composition and whether your PSA is designed for taking such gases. If the gases contains more amount of higher hydro carbons then it may plug the adsorbents and affects its efficiency then also the recovery of hydrogen may come down to such low values. In this case the Adsorbent has to be changed. Some may also regenerate the adsorbent by using Hot Nitrogen gas in-order to increase the recovery and remove the impurities. If so consult with your PSA licensor.
4. Generally the life of the PSA adsorbents are equivalent to the plant life as it is regenerated continuously by pressurizing and de pressurizing.
You can check with any of the above points and then decide for the adsorbent change over as it is not a simple job. one has to take shutdown of the unit and loading, unloading and start up takes a long time.
12/07/2010 A: keith bowers, B and B Consulting, kebowers47@gmail.com
You did not provide us with any data that could be used in formulating an answer to YOUR problem. PSA mol sieves CAN have a very long life (a decade or more) if they are suitable for the operating conditions and are not damaged by exposure to chemicals and conditions that either clog up the pores or cause collapse of the porous structure. It is VERY important you conform exactly to the design composition level of feed gas, pressure, and temperature specified by the licensor I recommend you discuss the issue in detail with the process licensor and vendor of the mol sieve load.