This is some conversation about questions that arose when I finally saw the O&M Manual. Not sure why it took the manual to ferret out some answers I though engineering could help me with BEFORE the unit arrived here.
Subject: Re: Feed Water Analysis
Author: Ionics rep
Date: 11/9/98 12:23 PM
Russ,
Sounds like you're talking about the O&M manual. I met with (Ionics engineer) today to discuss your questions. Please read on.
Regarding the question of feedwater conductivity beyond 20 uS (HP wants to go to 250 uS); although this is a special application, Ionics' will still need to get a handle on: ionic concentration variations, where the potential foulants are, how to treat for, and how to clean if need be. This can only be done by quantifying what the HP process is delivering as a feed to the EDI unit.
The Transformer is rated for 6 KVA, our max amperage is (*). If the process requires, we can replace the transformer with a larger unit. However we cannot tell exactly what the stack current will be in your application until we get under way and hit steady state. Please note the EDI process operates best when given the opportunity to hit steady state, long run times are desired with short down times. Have you quantified what the normal feed conductivity range during a 24 hr operation band will be?
>>The curve you sent me showing feed and product uS--will the unit I have in house be able to do that with the transformer provided? We don't need exact stack current-a qualitative yes or no would be good.
Assume it's at steady state. If we have to change the transformer, I'd bet there are other components that would need changeout too. The current draw is related to ionics removed. You must have some idea if ** amps is within an order of magnitude of what's needed.
I assume there will be short down times! Since we don't need 16 meg +-.1, but rather ANY meg above 1, are we really striving for an optimum, or will 'sufficient' suffice? Won't the EDI perform even if the feed goes from 5 uS to 200 uS over, say, 1/2 hour? The normal feed will range (for testing anyway) from 4 uS to 250 uS.
The EDI skid will require a feed pressure of 25-30 psig. It is in the EDI's best interests to keep this pressure as steady as possible. Can you install a PRV (set point = 30 psig) down stream of the ACF (activated carbon filter)? Variations in the supply pump suction pressure will vary discharge flows and pressure. This in effect will change the EDI dilute in, conc. in, conc. and electrode makeup flows and pressures. Stack differentials and even possibly cavitation of the product and conc. pumps could occur. Naturally these hydraulic variances will effect the EDI performance.
>>>A PRV or a regulator? The amount of pressure change would not be enough to go from positive to negative suction heads, but could easily swing beyond the 2-4 psi delta you require. We can have our contractor put in regulators if you can tell me where we need them. Mark it on your drawing for your next rev. product.
Please note the EDI ion exchange resin is in different states of regeneration throughout the length of the flow path. The end of the flow path being the most regenerated is the region responsible for weak ionic species removal and the front 30-50% responsible for strong ionic species removal. Modifications in system flow rate, change the point of removal in the EDI dilute flow path for the strongly and weakly ionized ionic compounds (would) basically effect each ionic species removal percentage (i.e. SiO2 could drop from 95% to 80% if the flow increased).
>>that's OK by me. I don't expect much SiO2. I'm only measuring conductivity. Am I concerned where in the stack my 1 meg is reached? I don't think so (yet).
Therefore we recommend maintaining steady state system flow rates.
>>OK. regulators. Tell me where to put them.
Although our membranes thrive in an acidic environment, we still need to place controls on the pH limits of the environment. The chemical feed system would be compatible with NaOH. Note the metering pump is strictly a manual controlled unit.
>>Unlike any other metering pump in existence today. Mygosh, how much would a pH controller have cost? Mind if I put one on? Got a concentrate port to sample from? What's the pressure at the input of the recirc pump (feed port)?
I will ask Li Zhang to review your question regarding balancing the feed pH. For now lets keep the pH > 1.0
>>Is there a pH meter on the unit? I can't find it. Good idea?
The stack differential pressure is controlled manually by varying the system feed pressure, feed flow, conc. makeup flow, conc. pump discharge pressure/flow, and product transfer pump discharge back pressure/flow. Generally once the makeup and system feed flow and pressure are set and at steady state the only two adjustments if necessary are through the Conc. pump discharge DV and Product Pump discharge back pressure control valve.
>>Is this how stack outlet pressure is controlled? By (manually) throttling the pump output?
11/13à Yes; but with the discharge back pressure control valve.
As stated in (Ionics engineer) proposal daily data sheets are required, usually this is when the operator will find a differential discrepancy. What we recommend is they finish the data set, correct the differential and take another data set an hour later.
>>By the time we find a discrepancy, is it too late? Sounds like you'd need a regulator on the system feed, none on flows.
>>From the manual: 'pressure of the demineralized stream is maintained at 2-4 psig higher than the pressure in the concentrate stream'. The concentrate stream and the product streams have pressure drops that change depending on the distance along the stack. Do these pressure drops track? That is, say that at the product input end the pressure is 30 psi; outlet is 10. Concentrate stream MUST be at 27 psi (+-1) inlet and 7 (+-1) outlet. That how it works? Or is it just outlet pressure that needs tight control--inlet is not a problem? That seems like a lot to ask, and gauges aren't even that good. How should this REALLY be done to free up a technician?
>>>I want to work with you on this to the extent that we don't need a technician to baby-sit it. I fully understand that in a RO-Primary application, this would be a snap. That's the opportunity here. Automate. Clean up the manual adjustments. I like doing controls and maybe we can put some pieces together (I have a spare pH meter/ controller and a few regulators, for example) to get the EDI marketable for this application. A proper working unit that doesn't need pressure adjustment is required. In the event of a pump failure, if damage can be done, shutdown should be immediate. You know, make it robust. I hope we can develop this together. I have numerous installation possibilities when we're done.
Thanks, russ
The questions were more-or-less answered when the unit was installed, the Manual was studied, and ** was cornered. Here are my notes from Nov. 12, 1998.
2. Is the limit of ** amps (**, power supply and/or transformer)) consistent with our desire to feed some chemistries at 250 uS?
11/12à The jury is still out. ** is unsure if testing done with 350 uS (!) water used special equipment or was just the standard. Ionics has offered a bigger transformer if needed; the controller can take ** amps, beyond the 100 gpm needs.
3. What is the NPSH requirement of the main pump? In a permanent installation, a proposed setup would be: 3Kgal tank, pump, carbon bed, EDI stack (no pump here--it's now in front of the carbon!), product pump, UV, product out to RO tank. The NPSH will vary as the tank level changes; and downstream, the pressure to the EDI. Do the concentrate side and product side pressures vary in sync with total feed pressure so that the differential membrane pressure requirement (2-4 lbs) can be met?
11/12à The stack needs 30 psi; this is where HP put in a regulator to absorb (existing) feed pump variations. Concentrate and product do not vay in sync. The stack resistance to flow is equal for product and concentrate. The drop across the stack from the feed water is matched by that of the brine recirc pump, at both ends (30 psi inlet and 0 psi outlet), with manual adjustments. With a back-pressure controller on the product out pump, variations in hydraulic resistance in the flow path (recirculate the water locally or pump to another building) are compensated.
4. The chemical feed may be a blessing for us. We expect our concentrate stream to be pretty thick with ions all the time, so we likely wouldn't use injection. But, the CIP process requires the pH to be above 1.5 or membrane damage will result. There's a possibility that the concentrate will see a lower pH. Can the injector be programmed to inject if that happens, rather than on conductivity? And perhaps, would it be better to inject the NaOH into the feed stream to neutralize the acid there, lower the conductivity (proton versus sodium) and more balance the EDI operation?
11/12à Rep says don't worry about the pH of 1 (not 1.5) limit. We'll look into pH control as we increase the contamination level to the EDI. Grab samples will be sent for pH measurement.
5. The membrane differential pressure--is it PLC controlled? Related to this, could someone put more time in on the Index so I can find the reference to MDP faster? (:>) Ahh, 5-7, #20. How can we keep 2-4 psi manually? Page 6-2 states that the 'pressure of the demineralized stream is maintained at 2-4 psig higher than the pressure in the concentrate stream'. I HOPE that's the PLC, not my vigilance.
11/12à The differential is manually adjusted. It's not difficult, just an unfortunately delicate adjustment to maintain.