RO Cleaning Cycle, when to do it

[dbill74]

We've got a small seawater RO desalination plant (1.6 MGPD) with 6 "trains" (set of filters, RO memebranes, pumps, etc). A recently completed project replaced 4 of these trains. There is some confusion about when to initiate the cleaning cycle of an individual train. Reviewing the plans, it appears the only way to measure/evaluate the amount of fouling on the membranes is to measure the pressure differnetial between the Feed and the Concentrate.

Membrane manufacturer's literature (SWC4-LD) indicates a maximum allowable membrane pressure drop of 10-15 psi. When asked about a maximum pressure drop when seven (7) membranes are in a single pressure vessel, the rsponse was up to 50 psi.

The system designer is telling us for the same seven membrane pressure vessel to go with 10-15 psi.

Prior to the the replacement project (used the same process), the plant operators were cleaning the RO membranes at 25 psi.

What would you recommend telling the operators?

My gut is telling me they should be cleaning at 25 psi still. Is this reasonable or am I off?

Thanks.

 

[bimr]

Not sure why you would not follow the recommendations of the system supplier. It would void a warranty if you have one.

RO cleaning frequency due to fouling will vary by site. A rough rule of thumb as to an acceptable cleaning frequency is once every 3 to 12 months. If you have to clean more than once a month, you should be able to justify further capital expenditures for improved RO pretreatment or a re-design of the RO operation. If the cleaning frequency is every one to three months, you may want to focus on improving the operation of your existing equipment but further capital expenditure may be harder to justify.

It is important to clean the membranes when they are only lightly fouled, not heavily fouled. Heavy fouling can impair the effectiveness of the cleaning chemical by impeding the penetration of the chemical deep into the foulant and in the flushing of the foulant out of the elements. If normalized membrane performance drops 30 to 50%, it may be impossible to fully restore the performance back to baseline conditions.

Cleaning operations performed at the extremes may result in a more effective cleaning, but can shorten the useful life of the membrane due to hydrolysis. To optimize the useful life of a membrane, it is recommended to use the least harshest cleaning solutions and minimize the contact time whenever possible.

 

[mk3223]

IMO, you may want to consider the total operation and maintenance costs in order to find an optimize pressure drop for the cleaning cycle of the RO system. The energy cost to operate the RO system will be up due to the higher pressure drop of the RO equipment. But, the maintenance could be lowered because of the longer cycle. It can be one of the system improvement project for your plant operation after some cycles.

 

[ashtree]

Bimr is absolutely correct about the cleaning when lightly fouled. This is normally easier and it will help maintain membrane life. Operating at high differential pressures increases the likelihood of premature membrane failure.

The 50psi quoted by your supplier is the maximum feed to concentrate differential pressure permitted for 7 membranes in a single pressure vessel. At least one membrane supplier says " But do not use 50psi as a target to indicate the need for cleaning."
Typically the recommended target is when pressures increase about 10-15% of clean pressure when normalised for temperature and salinity.

You have not said what your clean differential pressure is , so it is impossible to know whether 25psi is the correct figure. Many plants often run with 25psi differential from new, but this depends on many factors including , membrane used, recirculation rates (if used) and then is influenced by fouling rates.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[dbill74]

You have not said what your clean differential pressure is

That's because I don't know it, yet. One train was cleaned yesterday and I am hoping to get that information today.

I don't disagree with Bimr abuot cleaning when the membranes are only lightly fouled, in fact read his source prior to writing the OP.

you may want to consider the total operation and maintenance costs in order to find an optimize pressure drop for the cleaning cycle of the RO system.

That's the whole purpose of this exercise and my post. What I am wanting to get at is finding a good starting point.

If the sysetm designer and membrane manufacture were telling us similar figures, this would be a moot point. As it is, the "recommended" figures are not close. As an engineer I'm used to being conservative, but in this case I get the feeling the designer is being too conservative.

 

[ashtree]

If you have good records going back to when the plant was new it would be worth plotting out the differential and system pressure after every clean was done. Provided you know that the cleans have been done properly you will see a gradual rise in the pressures for the first few months of operations then a gradual flattening. Points along this flat part indicate the likely clean pressure that you are now trying to figure out.
If the plant is new or you dont have such data look at what your clean pressures are and use that as your initial base line.
Pressures can rise with, decreasing temperature, increasing salinity, increasing flow rates or changes in recovery in addition to fouling or scaling and you will need to correct for these changes.
Once you have done that a 10-15% increase in your differential pressure (feed to concentrate) or trans membrane pressures(for a constant flux rate and recovery) will be an indicator of the need to perform a clean. Cleaning at this point should be relatively quick and easy and you have not stresses the membranes mechanically from operating at high pressures.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[dfmorvan]

Do some flow normalization (there are spreadsheet available from membranes suppliers. When normalized flow decreases of 15% or norm pressure drop climbs 15 % it is time to clean.

 

[chief]

A pre -filter of diatomaceous earth/sand , will greatly extend the working life of your membranes and reduce the neccesity of cleaning.
Ensure a back flush arrangement is installed on the pre-filter, to facilitate its cleaning without the time-consuming job of dismantling.

Offshore Engineering&Design