By C.F. ‘Chubb’ Michaud, MWS
A few years back, I was invited to chair the somewhat recently created Commercial & Industrial (C&I) Section of the Water Quality Association (WQA). The purpose was to organize and develop the C&I activities, in order to differentiate their concepts and practices from the predominantly residential mindset that defined WQA at the time. Our first order of business was to define where residential left off and C&I began. Two years later, that task was yet to be concluded.
A few of the old-timers on the section committee (who worked for large water treatment companies that divided their sales teams according to Residential and C&I sales) professed that, “if the feed pipe was two inches or larger, it was industrial. If it was one inch or less, it was residential.” Commercial was considered a smaller version of industrial. Another old-timer with a similar organization disagreed with that and maintained that it was solely dependent on the tank size.
Up to 14-inch was a residential system while 48 inches and over was industrial. Systems that used more than one technology (i.e., a carbon filter and a softener in front of an RO) were indicative of an industrial design. The real question was and still is: Does size really matter? Or is it based on something else?
The need defines the term
Let’s look at a couple of examples. There are three establishments on the same block of Watertown, USA. One is a boarding house with four tenants and the landlord. One is a drive-through car wash and a third is a pharmaceutical supply house that produces a purified water for injection (WFI). All are on a city water supply that runs 20-gpg total hardness (TH) and 550-ppm TDS. All claim they need softened water.
The boarding house includes a kitchen and laundry facilities, in addition to five bathrooms. Their peak water flow is deemed to be 10 gpm and they average 600 gallons per day. The semi-automatic car wash sprays, brushes, rinses and blow dries, and the washed vehicles are toweled down by hand. On warm and sunny days, the residual water drops might dry down and show hard-water spots, hence, the need for a water softener. The sustained water demand by the line is 10 gpm and they operate 12 hours a day using about 3,500 gallons of water. The pharmaceutical house operates a vacuum distillation system to produce WFI. Feed water must be free of scale-forming total hardness. The feed rate to the still is 10 gpm and they produce 35,000 liters (9,250 gallons) per day (16 hours) with 20 percent blow down.
So here we have three systems, all having the same flow demand: one residential, one commercial and one industrial. We shall call them R, C and I. All are seeking softened water. Do we have three identical systems and if not, how are they different?
The first consideration here is to evaluate the burn rate. How many grains of hardness must be removed per day and how many cubic feet of resin will be required. The boarding house (R) uses 600 gpd. At 20 gpg, it has to remove 12,000 gr/day. Based on a typical 8.0-lb/cu ft-capacity of 24,000 grains, R’s burn rate is 0.5 cu ft/day. The car wash (C) uses 3,500 gpd and must remove 70,000 grains of total hardness. Again, based on 24 Kgr/cu ft, C’s burn rate is 2.92 cu ft/day (we will call that 3.0 cu ft). Our distiller (I) has to produce 35,000 liters (9,250 gallons) of WFI and with the 20-percent blow down, it must feed 9,250/0.8 = 11,560 gpd and remove 231,200 grains. I’s burn rate is 9.63 cu ft/day.
Monthly salt consumption is estimated at:
Brine tank size Salt capacity
R: 0.5 cu ft/day x 8 lbs/cu ft x 30 days/mo = 120 lbs/mo 18 x 33 350 lbs
C: 3.0 cu ft/day x 8 lbs/cu ft x 30 days/mo = 720 lbs/mo 24 x 60 1,000 lbs
I: 9.63 cu ft/day x 8 lbs/cu ft x 21.6* days/mo = 1,665 lbs/mo 39 x 48 2,200 lbs
*21.6 is the average number of working days per month less weekends and holidays.
So far, we have three pretty similar systems except for the size of the brine tank. The choice of brine tank was to provide a delivery service to refill the salt no more than once a month. What else is different?
To assess the path…
You have to know where you will start and where you will end. Before we go ahead and program the control valve, let’s look at the quality requirements for the three systems. The landlady at R says that the cleaning lady complained that it was difficult to clean the shower doors and dishes with hard water. Also, the linens were a little scratchy. WQA maintains that softening to below 1.0-gpg total hardness eliminates these issues, so we shoot for a typical softener performance system that is adequate to the need. To handle the 10-gpm flow requirement, we would use a 14×65-inch pressure vessel with 3.0 cubic feet of 8.0-percent cation resin and plug in a 500-gallon reserve. The R unit will regenerate every five to six days, but because water usage on traveling holidays and party weekends will introduce a wide variable in flow demand, we will add a meter and make this a DIR system. There is ample downtime to regenerate any night of the week so there is no need to double up with a twin alternating unit.
The crew down at the car wash complained that on sunny and breezy days, water spotting would leave hard scale deposits on the car finish which required a lot of extra elbow grease to clean up. Hardness leakage with an 8.0-lb/cu ft salt dose is about 8.0-10 ppm (see inset). The system supplier stated that softening to less than 2.0 ppm total hardness would all but eliminate hard-water spotting. In checking the leakage charts, we find that to take our 550 TDS water down to < 2.0 ppm total hardness would require brining at a level of 22.5 lbs/cu ft. Although the capacity will go up considerably, that’s just an awful lot of salt.
Maybe it should be run with up-flow regeneration. A rule of thumb with true up-flow regeneration is you get the same capacity but with 1/10th the leakage. So, again, we choose a 14×65-inch vessel with 3.0 cu ft of 8.0-percent resin and up-flow regeneration. On those long summer days with longer daylight hours, the car wash often opts to remain open for a few extra hours. That means that their softener capacity is apt to be run to the ragged edge. Shutting down the line to wait for a regeneration is out of the question. We will meter this system and make it twin-alternating so there is a continuous flow of soft water.
Hardness leakage is not really a result of too fast a flowrate (blow by) or the approaching end point of the run (breakthrough). Leakage is the low-level hardness bleed that occurs during the service run and is caused by the low hardness, softened water passing over resin that contains residual hardness. This is particularly evident for down-flow regeneration, which deposits a fair amount of hardness ions swept from the top of the bed onto the relatively clean bottom portion of the resin bed. The bottom of the bed is partially regenerated by the soft water during service (an equilibrium thing) and shows up in the effluent. The higher the salt dose, the lower the residual hardness and the lower the leakage. Counter-flow (up-flow) regeneration puts the strongest and cleanest brine through the least exhausted part of the bed and produces very low levels of leakage.
The WFI maker presents a different set of conditions. The feed water will be heated, which converts the soluble bicarbonate salts into potentially insoluble carbonate salts. In addition, the vacuum still will concentrate any residual hardness by a factor of five or more. Even 1.0-ppm leakage can eventually shut down the operation due to scaling and plugging of the narrow flow channels within the still. That means lost production, costly repairs and downtime. In addition, WFI must be periodically heat-disinfected and can only be stored for a few days in bulk before it is dumped. They will need a more bulletproof unit with not only a high degree of reliability but a higher degree of performance as well. The still manufacturer has set a specification that the residual hardness in the feed water be less than 0.2 ppm. What considerations will you make to differentiate the level of performance required for this application?
With industrial applications, the question is not one of pipe size or flowrate. It really has to do with the level of performance and the reliability to continuously produce that given level of performance. We must ask ourselves: “What are the consequences of failure?” With unit R, the consequences of failure are some water spots and a complaining housekeeper. It’s a short-term situation that we can fix later in the week. With unit C, production time is at stake but it still comes down to: “Gee…it would be nice if we didn’t have to work so hard to wipe down these cars.” No permanent injuries or damage will be incurred. To make the system work, however, we reconfigured the regeneration mode and backed it up with a metered twin-tank system, which gives the operation a day or so to get things fixed. No on-site storage is needed.
If unit I fails, there can be some real damage and heavy expenses incurred, in addition to the complete loss of production. WFI sells over-the-counter for about $10/liter, giving the daily production value of around $350,000. With a burn rate of 9.6-cu ft/16-hr work day, a 3.0-cu ft unit only gives five hours of production. We should provide at least a full day of backup with a 9.6-cu ft unit. Up-flow regeneration is a given but in this case, for consistency, we have to design the system to hold the bed in place during regeneration. Now it becomes a packed bed with an inert resin packing at the top. We have eliminated the backwash step but must provide prefiltration to keep the packed bed as clean as possible. With more resin, we need a larger tank. The choice has to give us at least 4.0 gpm/sq ft for good distribution. A 2.5-sq ft tank is 21 inches in diameter (21 x 62) and (with under-bedding) only holds 9.0 cu ft (with head space for inert resin and inlet distribution).
Is 8.0 lbs/cu ft enough salt to insure 0.2-ppm leakage? Perhaps not. We should increase that to at least 12 lbs/cu ft. Our capacity will jump to 30-Kgr/cu ft and our burn rate drops to 7.7 cu ft/day. This (fortunately) provides us with an excess capacity of 15 percent, which is exactly the engineering factor we would apply to this system. An engineering factor is an intentional downgrade of the system’s capacity in order to reflect the real-time percentage capacity loss over time. This allows for a 15-percent loss of capacity over a period of three to four years. Why not consider a different type of resin, such as uniform beads, to improve the efficiency of the softening process?
Should we install a polisher…just in case? Set it to regenerate once a week? Good idea. While we’re at it, let’s add a battery backup for good measure. That’s just in case a power outage messes with the valve memory.
All of these considerations contribute to a warm and fuzzy feeling as to the performance of producing 10 gpm for 16 hours on 20-grain water softened to 0.2 gpg. The twin alternating gives us a full-day backup and a polisher of equal size and regeneration mode seals the deal. This unit will work.
Are you sure?
How do we know that? We can’t be 100-percent sure, so we monitor the system by performing a low-level hardness leakage test every four hours or install a hardness monitor with an alarm that gives continuous data and automatically switches the tanks if the alarm condition goes above a set-point for more than 10 minutes. Where do you install the sensor? It would go after the primary and in front of the polisher.
What if we run out of salt? Having a one-month inventory makes things easy to remember but daily checking of the brine tank level has to be routine. Increasing our salt level to 12 lbs/cu ft (times 9.0 cu ft) = 108 lbs/day or 2,330 lbs/mo. This will slightly exceed our original 2,200-lb brine tank, so we set aside four bags of salt and add one every Monday morning. We also put a low-salt alarm on the brine tank that gives a visual (flag or light) plus and audible signal when the salt is running low.
What if the eductor gets plugged and the regeneration fails? The quality monitor will tell you in a hurry that the system failed to regenerate. A conductivity monitor in the brine line or waste line will tell us that the TDS spiked for a given length of time, which would indicate that the brine was drawn through during the regeneration cycle.
Assessing the consequences of failure may seem arbitrary, particularly to the end user who has to pay for it. Selling an industrial user with a real industrial need requires knowledge of the application requirements as well as a logical and well-organized presentation of the equipment being offered, including why it is designed the way it is. To many users, a softener produces zero hardness but indeed, it does not. A good industrial design will hedge on many of the things that could go wrong but machines still have their limitations. Most applications will require the overview of a human operator who also understands what the machine does and doesn’t do. So the final element of a good industrial design has to include training and an operations manual and follow-up.
About the author
C.F. ‘Chubb’ Michaud, Master Water Specialist is the Technical Director and CEO of Systematix Company of Buena Park, CA, which he founded in 1982. He has served as chair of several sections, committees and task forces within WQA, is a Past Director and Governor of WQA and currently serves on the PWQA Board, chairing the Technical and Education Committees. Michaud is a proud member of both the WQA and PWQA Halls of Fame, has been honored with the WQA Award of Merit and is a two-time past recipient of the PWQA Robert Gans Award. He can be reached at (714) 522-5453 or via email at AskChubb@aol.com.