Wayne Robey with editing by Zhahai Stewart

ABSTRACT. After considering various alternatives and options, a filter for filtering water to make it potable is presented. It consists of an optional prefilter, a Katadyn ceramic filter, and an optional activated carbon post filter. There are several options for forcing water through the filters. All are based on the principle that a small pressure for a long time is more efficient than a high pressure for a short time. These include elevating a tank to provide pressure due to gravity, using water that is already elevated, amplifying that pressure with a hydraulic ram, and using an electric pump.

I have been exploring several low cost options for purifying drinking water. The lowest initial cost and least energy efficient method is to heat the water to boiling. Heating it to 149 degrees F for 6 minutes (pasteurization) kills the major pathogenic organisms including virus. Energy requirements can be reduced to a reasonable level if a very good heat exchanger is used to heat the raw water with the pasteurized water. Surface water should be prefiltered with the 1 micron prefilter discussed later and will benefit from the activated carbon filter discussed later. The main problem with this is that a thermostatic valve is required to regulate the water flow. Years of experience with these valves clearly show that affordable valves are far from hippie proof. A write up on this method appears in Home Power magazine #52. If you wish to construct such a device, please consult with me.

I believe that filtration will be the most generally useful method in the Rainbow environment. The systems I present below are intended to make filtering affordable to the average and large kitchen. Further investigation revealed several general principles: Contaminants in water vary so widely that definitive answers to the question of what filters are best or their expected life cannot be given. A completely hippie proof system is not possible. Hippie proof withdrawal of water is possible, installation is close, but cleaning the filter I recommend is not. This does not mean that unsafe water may be produced, but that poor cleaning practices will result in poor filter life. Anyone cleaning the filter should read the section on cleaning. Some have ruined a good filter by totally disregarding the cleaning instructions. The best system configuration will vary greatly with the designed capacity and local conditions. It is intended to be kitchen sized as a minimum but is suitable for an individual or to provide over 2 GPM continuously when used with hydraulic ram, electric or other pump. It is easy to setup and intended for a camp of short duration such as a rainbow gathering but not for a group that is on the move. I can provide more detailed drawings by US mail on request, but unless you have the parts and facilities on hand, you will probably want to buy at least part of this ready-made. My mailing address is:

Wayne Robey, 215 S. 6th St., Lafayette, Indiana 47901.

PARTICULATE FILTER SELECTION

As the filter becomes finer, the flow rate decreases rapidly. Removal of most virus by filtration is not practical, if this capacity is desired, it would be desirable to filter water through a good prefilter into a storage tank, treat the water with iodine or chlorine, and after sitting long enough pass it through a high capacity activated carbon filter. If the carbon filter is not cleanable, it should have a suitable prefilter to keep it from clogging. In the North American wilderness, virus is not generally considered to be a problem but a large gathering is not wilderness. The details of this process are not included in this paper. For complete removal of microorganisms the generally accepted requirement is 0.2 micron though 0.4 micron works nearly as well. E-coli is one organism that could occasionally slip through a 0.4 micron membrane filter but that is not disastrous. While cryptosporidia and giardia vary, some believe that they are readily removed with a 2 micron filter but as protozoa divide, a finer filter is required. One thing to watch here is that some 0.4 micron nominal filters will pass a significant amount of 10-micron particles and are only useful as prefilters. In general a non-cleanable filter will pass more water initially but have a much shorter life than one that can be cleaned many times by removing the contaminated surface. Water quality from a questionable or contaminated source varies greatly and cannot be predicted with any confidence.

Specifications for typical or maximum life in terms of gallons of water filtered are meaningless. Since there is no standard water for rating filters and particle size distributions vary greatly, these advertised lives cannot even be used reliably to compare filters. I have seen the cleaning frequency change by a factor of 5 from the same source over the period of a week.

Water from the Wabash river reduced the flow of a Katadyn 1040 filter to 60% of the initial value after 2 gallons; but at the 1996 North American Gathering, using spring water and a 1 micron prefilter, 160 gallons of water could be produced before clogging to that degree. Deep well water will usually produce more water before clogging than this, while pure water will never clog the filter. At the 1998 Ocala gathering, fine particulates penetrated deeply into the Katadyn ceramic. For these reasons, a great deal of surplus filter life should be available.

For general use, I settled on two of the Katadyn ceramic filters: the LP-K700 (also referred to as article #1700) and the #4 (also referred to as article #1040). Both have a nominal pore size of 0.2 microns and are well established as reliably producing safe water. The main differences between them are shown below. Both have advantages and disadvantages. The main advantage of the 1700 is the high throughput to size ratio. The main advantage of the 1040 is long life under favorable conditions and better filtration of particles smaller than 0.2 microns. The flow rate as a function of pressure is linear over the useful range. The new flow rate is not fully restored by light cleaning due to the fact that the ceramic contains some large pores that increase the initial effective surface. In addition, clay, smoke, carbon black, and possibly other inanimate particles exist as a continuous distribution of sizes over the 0.05 to 0.3 micron range. Virus also exists in this range but clay is the most common problem. Some of these will penetrate the ceramic to great depths. This is a greater problem with surface water than spring water and much worse if the water is muddy due to a recent rain than if it is clear. This should be kept in mind when selecting a gathering site. This has more effect on the #1040 than on the #1700 because of its greater life. On the other hand, lake water filtered through a new #1700 at the 1998 Ocala gathering was not as pleasing as that filtered through the #1040 due to the greater ceramic depth of the #1040. If the water has been favorable, the #1040 filter approaches the performance of the #1700 as it ages. The chart below shows the conversion between clean filter, instantaneous, and average flow. The cleaning point is the instantaneous flow when the filter is taken down for cleaning. The quantity filtered is the water filtered in arbitrary units before the filter needs to be cleaned again. (The conversion from these arbitrary units to gallons depends on the water quality.) Since each cleaning removes about the same amount of filter ceramic, this number is inversely proportional to filter life. It appears to me that for system sizing, it is wise to plan on cleaning the 1040 filter when the flow is half the initial flow but in use, it is wise to clean less frequently when there is no shortage of filtered water.

Quantity filtered in relative terms, only for comparison within this table. 1.0 = unit #1040 cleaned when flow diminishes to 1/2 intitial. Flow in gallons per 24 hr day @ 1 psi.

PREFILTERING

Under many conditions, a 1-micron prefilter is useful. It always greatly reduces the cleaning frequency. It will generally reduce the lifetime operating cost of the #1700. It is not clear that this will be true for the #1040 due to it's long life and the life time cost of the prefilters. The first suggestion for a prefilter is a 1 micron nominal pleated polyester element with as much surface area as possible. This has provided good life over a wide range of water conditions while greatly extending the life of the ceramic filter. I suggest a 9 3/4" double end cartridge with 6 sq ft of media and sold by Mc Master-Carr as part 45235K43 for $5.85. I do not have extensive experience with this element and welcome your experience as to its cleanability. Often 2 cleanings is the practical limit since capacity is reduced with each cleaning. In order to avoid making a special housing for these, I have used a common household filter holder with the input and output connections identical for backwashing. To prolong prefilter life, installations which use more than 2 ceramic filters in parallel should consider using prefilters in parallel.

The holder for the #1700 is based on 2" schedule 40 PVC parts, uses a friction fitting test plug and is the easiest of the three to make. The filter is supported on both ends for ruggedness.

The best holder for the #1040 is based on the uncommon 2 1/2" schedule 40 fittings. I have been unable to find a suitable 2 1/2" test plug and decided to use an entirely different design which requires a lathe. Cleaning the filter is quickest with this design. The filter element is supported on both ends for ruggedness.

The third holder can be used for either element and is based on 3" thin wall PVC parts. It is not much bigger than the second, uses much less PVC, but is not rugged. There is no pressure rating on the parts, some of which seem to be poorly made. This would be a good choice for home use at pressures not exceeding 5 PSI. I suggest that some means be provided to support the bottom of the filter elements during transport. A concentric prefilter could be built into this housing (probably using polyester felt) but would be clumsy due to the frequency of cleaning the inner filter.

All of these holders require a special nut (for the 1040 element) or special tube connector (for any filter element) which I can supply.

The friction fit test plug is used rather than a pipe thread fitting because I do not believe the pipe thread would have the durability required for the hundreds of cleanings required over the life of the housing.

To keep any installation of more than one element manageable, it is important to rigidly mount each filter holder as well an the input and output manifolds. Additional detailed information as well as the finished holders are available. I believe that obtaining the hardware from me is more practical than making your own unless you prefer a different design or prefer to use different materials.

CHEMICAL REMOVAL THOUGHTS AND OPTIONS

It may also be desirable to remove various chemicals but it may not be worth the trouble if precautions are taken in selecting the water source. Activated carbon and selective ion exchange resins can do this but have a variety of limitations so it is important to understand what you want to remove. In addition, as the filters age they give no indication of their effectiveness. Overall, this cannot be hippie proof.

The most common problem with surface water will indicated by poor taste due to various metabolites. These are also the easiest to detect and remove but minerals can also have an undesirable taste and will not be removed by activated carbon. Agricultural runoff is a potential source of herbicides and pesticides as well as nitrates. Toxic metals can result from mining or other industrial activity. Chemical contamination can result from small-scale activity and be unsuspected. In some areas arsenic occurs naturally.

The first thing to realize is that as these filters become saturated with a specific contaminant, they let more and more pass through. If you can't detect it and don't know how much is in the feed water, you can't be certain that any filter is doing its job. While free chlorine and many odors are easily detected and removed, many organic chemicals require varying degrees of long contact with high activity activated carbon and are not easily detected. As the system flow goes down contact time (and effectiveness of the activated carbon) increased. This is another advantage this system has over those that are pumped to a high pressure by an impatient operator to get high flow for a short time.

To prevent growth of organisms in an activated charcoal filter, other organic materials must be excluded by placing it after the bacterial particle filter, or it must be cleaned frequently.

One activated carbon filter worthy of consideration because it combines small size and unusually low flow resistance with very small uniform pore size and large surface area is a polyolefin bonded powdered activated carbon briquette with a surface area of 2,200,000 sq ft and a pressure drop of 0.9 PSI at 1 GPM sold by Cole-Parmer as cat # E-01508-93 for $26. The Anetek CDC-10 is an alternative (1.3 PSI @ 1 GPM). An alternate approach adds the ability to remove some fluoride, lead, cadmium, mercury, and arsenic with hydroxyapatite charcoal. The granular blend, called R1022, is a more general purpose activated carbon and can be cleaned so it can be used as the first prefilter but it requires more attention than the above briquette. It is available in 80 cubic inch packages from The Rockland Corp. as stock number 8111 for $18.20. Each package should be able to treat 1/4 GPM. See the section on using this media. I think everyone using city water should have this at home.

Selective ion exchange resins are useful to remove metals such as lead, cadmium and mercury which may be introduced by mining or pesticide use. One problem with these is that they have a poor shelf life, requiring annual replacement. Another problem is that they can support growth of microorganisms, they also require long contact time similar to the requirements of activated carbon. This makes these materials most suitable for applications requiring small amounts of water regularly where a known problem exists.

PUMPS

Since the energy required to filter water is proportional to the pressure, it is desirable to use a low pressure for a long time rather than a high pressure for a short time. I can think of three reasonable ways of producing moderate pressure for a long time:

• Continuously operated pump powered by motor, human or other power source.
• Pressure developed by gravity from water stored above the filter outlet.
• Gravity developed pressure amplified by a hydraulic ram.

The first case, continuous pumping, is great if you have electricity or the filtration system is large enough to keep a person pumping continuously. A hand operated piston force pump is efficient and easy to use but is not commercially available. Would someone like to make some? For an electric pump, a diaphragm pump operating near its maximum pressure is efficient and reliable.

The two that seem the most suitable are made by Flojet and are now sold by J.C. Whitney. Stock # 81JZ4553W costs $39.95 and is advertised as pumping 1.1 gpm @ 35 psi while using 3.5 amps @ 12 volts. Stock # 72JZ8421A costs $49.95 and is advertised as pumping 3.3 gpm @ 35 psi while using 7.0 amps @ 12 volts and could be suitable at 6 volts pumping 1.6 gpm. J. C. Whitney is not known for accuracy and I have not checked these specs. The best small diaphragm pump is the Shurflo 8009-541-236 (COST ~$80) which is a good match for two parallel #1040 elements when operated at 12 volts. Specs are 0.75 GPM @ 30 PSI & 2.8A, 5000 Hr brush life; 0.65 GPM @ 50PSI & 3.5A; 60 PSI max.

In the second case, gravity feed, if the water is naturally available at the desired height, connecting a pipe to it is all that is needed but this will usually be the case only when small amounts of water are needed or when the site is chosen to suit the water source. When only small amounts of water are needed it can be carried up to an elevated reservoir and allowed to flow back to the filter by gravity, pumped up by hand if you have a suitable pump, or a reservoir such as a 5 gallon bucket or dromedary bag can be hung from a tree by means of a pulley and rope, the water poured in at ground level and hoisted to the desired height. I don't like this solution because I am worried that someone will drop the bucket. For example if a height difference between the middle of the input reservoir and top of the receiving reservoir is only 10 feet and a single 1040 filter is used with the suggested cleaning schedule and a 20% initial loss of capacity is assumed, 10*0.433*13.85*(1-.2)= 48 gallons per day or 2 gal/hour could be produced. If the bucket is used, the filters and tubing can be stored in it. Whenever a reservoir that may run dry is used, at least one of the filters should have an air vent to remove the air in the lines when it is filled.

The third solution, a hydraulic ram, is very nice for taking water from a mountain spring or stream or spring but needs to be adapted to the site, requires some care in installation and an abundant source (which is not unusual). The major components are reverse flow input screen, (a 20 - 30 mesh screen seems sufficient to keep the valves working smoothly), supply tube if needed, stand pipe if input tube is used, rigid drive pipe, overflow valve, check valve and air tank to cushion the output. While I do not have detailed data and have not done any testing, it is reasonable to expect a cheap but well installed hydraulic ram to increase the total head by 6 to 8 times the head seen by the ram with an efficiency of 50%, greatly reducing the number of filter elements required. For example a 2" ram set for maximum production and having a 10 foot drop from a 30 gpm source through a 40 foot drive pipe could deliver 2.3 gpm or 3300 gpd at 30 psi (at the pump output) which is sufficient for distribution to a small network or at least six #1040 elements in parallel. A single 1" ram using 4.5 gpm under the same conditions would produce about 420 gpd using a single #4 filter under reasonable operating conditions (280 gallons for a 16-hour day).

Here are some design ideas for hydraulic ram pumps: The supply tube and stand pipe should be as large as practical in order to deliver the available head to the pump, the drive pipe length should be 3 to 6 times the head. For input flows near the minimum, the drive pipe should be near the maximum to store more energy in the pipe, for flows near the maximum the drive pipe should be short to reduce frictional losses. The drive pipe must be rigid, the ideal material for the drive pipe is steel but it is most commonly schedule 40 PVC. Some have thought that PVC should be avoided due to the environmental impact of producing and disposing of it. I think this is well founded but the alternatives must be considered. A PVC drive pipe would be the largest use of PVC by a large measure but the alternatives are steel or a radially glass reinforced plastic. The overflow and check valves are somewhat matched because they must work together. The trip point on the overflow valve (and resulting pump output) can be adjusted by adding or removing weight from the poppet. The check valve should have an opening backpressure of around 0.5 psi. Efficient operation requires the pump to be rigidly mounted. Clamping the check valve outlet to a bracket which is mounted to a couple of large rocks or stakes would work. There are several choices for the local storage tank which must be closely coupled to the check valve. As the number of filter elements being fed increases, the minimum size of this tank increases. In all cases a vertical 4" pipe of suitable length can be used. It needs a port on top to inject air (a tire valve and bicycle pump is good). Since air is dissolved in the pressurized water, occasional air additions will be required. This can be avoided by installing a flexible bladder in the tank. A suitably sized innertube could be put into the tank. If used, it must be folded, producing stress points. To prevent problems a large tube with a small enough amount of air to prevent excess stress at the folds is required. For short-term use this is probably not worthwhile but for a permanent installation it is a real convenience.

For pumps driving 1 - 2 filter elements, a tank can be made from a 2" pipe the bladder designed for the purpose available for $24 from the Ram Co. can be used. The specialized parts shown below are available from the Ram Co., 247 Liama Lane, Lowesville, VA 22967, 800-227-8511. The PVC check valve below may work better than the brass one, the poppet tees are overpriced and available locally. The maximum fall into the pump is 15' and the maximum practical lift above the pump outlet is 11 times the input fall.

The complete pump prices may have changed. Also it is not certain what air tank is used. The tanks with the Ram Co. bladder are too small for systems using more than 2 filter elements and are a bit questionable for the 2 element case.

SIZING EXAMPLE

First decide how much water you want. Lets say you want 50 gallons over 12 hours or perhaps in a 24 hour day but only attended for 12 hours, which is reasonable for a medium kitchen without many neighbors. Now lets say you want long term life for the filter in any condition rather than high immediate output. (You could rework this for good output with clean water input.) My assumption means that you want to use the #1040 element. If you look at the chart or table, you see that if you clean the filter when the flow reaches half the new clean filter value and allow for another 20% loss as explained in the article, the minimum instantaneous flow will be 3.7 gallons/12 hr day at 1 psi or 5.6 gallons averaged over a cleaning cycle.

If the water is clean and the pressure low, the time between cleanings will be long so the 3.7 gallons is used, realizing that you will usually have more. Then you can guarantee the 50 gallons from a single #1040 element if the pressure is 13.5 psi which you can get from a 31 foot fall. This is reaching the practical limit of what a manually fed system could achieve so you may want to consider the options:

1. Accept this fall requirement. Your comfort with this may depend on who is using it and the terrain.
2. Require less peak water production. You always have the option of cleaning the filter early to meet the occasional peak demand.
3. Change to a #1700 element and get the same output with a 20' fall.
4. Put two #1040 elements in parallel and get the same output with a 15.5' fall.
5. Operate more hours per day. Using a large receiving tank (in this example 32 gallons is sufficient) with continuous input from tank or spring, operating time can be increased to 24 hrs/day and peak demands leveled out.

Then the average flow in my table can be used with confidence provided the filter does not require cleaning overnight. Then the 50 gallons per day is obtained with a head of 2.3*50/13.85/0.8= 10.3 feet which is three times better than the initial design above. It is also much more convenient to manage one large tank than many 5 gallon buckets (though you could put a few buckets together - in series for the output or in parallel for the input).

CLEANING THE FILTERS

The sole advantage of a cleanable filter is longer life and reduced operating cost. Wasteful cleaning will eliminate this advantage. Katadyn supplies a cleaning tool they call a brush but is really a rigid (curved to fit the filter) very course abrasive which will take off a large amount of ceramic. If this tool is used, the advantage of a cleanable filter is largely lost. I do not know why they have provided this tool for these filter elements, but the approaches they used in some versions of their pocket filter are much better. I estimate an increase of life of five times or more when using the method I describe here as compared to the Katadyn "brush". (I have used water pressures of 30 PSI or less in my work while the Katadyn KFT filter can operate as high as 112 PSI which might drive the contamination slightly deeper, so my method could require modification when used with the KFT filters under some water conditions. Some users report that my method has worked well on their KFT.)

I estimate 75 cleanings for the #1700 and 300 for the #1040 element when Katadyn's recommended end of life point is used, but this will vary somewhat depending on the nature of the contamination.

When to clean the ceramic filter. With a continuously pressurized or gravity powered system never clean the filter if you are producing sufficient water. If you are not, measure the flow and compare it to the flow graph. One approach is to match the measured flow to the instantaneous flow on the graph and read the average flow. If these numbers are lower than you would like, clean the filter.

Remove filter from housing if necessary. Ceramic filters in the home made housings are best cleaned by removal. Disconnect the outlet tube; insert the protective plug to maintain the cleanliness of the element's interior; remove the filter from the housing; install the protective cap; immerse the element in water; and clean. Filters in 2.5" housings need not be dismounted for cleaning.

Brushing. If the element is not discolored, if it is shiny, or if it is slimy the first step is to brush it with a bristle brush. The bristles should be similar to those in a medium softness flat-bottomed toothbrush or a very short bristled paintbrush. This is to remove the surface slime. If it does not remove the slime it is too soft and if it removes a noticeable amount of ceramic, it is too stiff. Do not use a stiff scrub brush because it will gouge the soft ceramic. If a suitable brush is not available, a fine nylon scouring pad with little abrasive may be used. See the parts list below.

In every case, gently brush the ceramic as uniformly as possible with an abrasive pad. The common Scotch Brite abrasive pads used for cleaning kitchenware will do but Niagra light-duty scouring pads are easier to use for normal cleaning because they are less abrasive. If there is a dark deposit on the filter, brush it until the easily removed discoloration is removed. If the surface is not darkly colored it is probably covered with slimy microbes. In this case it is hard to know how much cleaning to do. There is no point to an incomplete cleaning but excessive cleaning is a complete waste of the ceramic, so start with slight cleaning with the abrasive.

The best indication of the amount of ceramic removed during cleaning is the color of the wash water. The best check on the quality of the cleaning is the flow rate after the element is reinstalled. Make the measurement after the filter has passed 1 gallon of water to insure that the air has been purged. If the flow is less than expected, clean more thoroughly next time to se if that is helpful. Deep contamination will reduce the flow and not be removed by normal cleaning, so if a little more cleaning has no effect on flow, take that as the clean filter value for future reference.

Reinstall. Rinse the neck of the element; remove the protective cap; insert the element in the housing, rinse the neck; install the nut or tube fitting on the filter element (IMPORTANT - tighten the nut or cap, but don't twist the filter element with much force -- the ceramic is not very strong); remove the protective plug; purge the air from the filter (optional); allow some water to flow through the filter; and install the outlet tube.

Cleaning the prefilter. The 1-micron pleated media filter is readily cleaned of large particles but not necessarily of small ones or of biological slime. To attempt to clean the filter: connect incoming water to the outlet and connect the inlet to a drain; introduce air into the filter housing till it is about 2/3 full of water; invert; turn on the water and shake. (Clearly the prefilter housing is connected by flexible tubing and easily moved. The other filter housings are best held rigidly.) If this cannot be done, the next best method is to remove the element from its housing and agitate it along its axis while submerged. If these filters are to be cost effective, they must be effectively cleaned.

Activated Charcoal. The powdered AC filters must be placed after the bacterial filter to prevent clogging, it should be disinfected by passing 3% hydrogen peroxide through it before using if it has been unused for over a week.

The granulated AC filters should be constructed so they can be disassembled for cleaning. The media can be poured into hot water and simmered for at least 20 minutes. This provides partial activation as well as removing particulates.

MORE USEFUL PARTS

Pressure gauges. In normal use pressure should not exceed 70% of full scale.

60 psi range, 1 psi marks, 2% accuracy, 1.5" dia, part 3847K2, $7.21 from McM

30 psi range, 0.5 psi marks, 2% acc, 1.5" dia, part 3846K3, $10.53 McM

15 psi range, 0.5 psi marks, cal @ 5 psi, 2" dia, $4.00 WR

Flexible PVC tubing. Easy to use for short lengths such as internal plumbing and running to a nearby container, can pinch off but lower pressure tubing is more durable when pinched off.

Silicone rubber, Durometer

2", 0.066 wall, 80 psi, polyester reinforced, lays flat until pressurized, easier than rigid tubing to carry 5295K35 $.67/ft < 100 ft, $.55/ft >95 ft, mult of 5' McM

3", 0.078 wall, 70 psi, 5295K39 $1.12/ft < 100 ft, $.93/ft >95 ft, mult of 5' McM

1 micron pleated polyester prefilter, 6 sq ft area in a 9 3/4" x 2 3/4" dia cartridge 45235K43 $5.85 McM

Powdered activated carbon briquette, ppolyolefin bonded with 220000 sq ft effective surface E-01508-93 $26 CP

Katadyn 1040 element $90 post paid in USA Leh (Better prices may be available, contact me for current info.)

Katadyn 1700 element $75 post paid in USA Leh (Better prices may be available, contact me for current info.)

PVC holder for 1040 element from 2.5" pipe, includes air vent and 6" outlet tube with pinch off clamp, element mounting nut, press on cleaning cap, mini wash bottle $20 WR

For screw on quick connect mounting adapter for more secure outlet hose connection especially with outlet pressure above 5 PSI, specify 1/4 or 3/8" hose connection add $3 WR

Nut only for your own holder $3 WR

Cap with 3/8" FPT outlet only, for your own holder (mounts filter element to holder and provides 3/8" FPT outlet) $6 WR

PVC holder for 1700 element from 2" pipe, includes air vent and 6" inlet & outlet tube with pinch off clamp, screw on quick connect mounting cap (provides quick connection to 1/4 or 3/8 hose (specify which)) & cleaning cap, mini wash bottle $12 WR

Cap with 3/8" FPT outlet only, for your own holder $6 WR

Multi element assemblies using these holders and pressure gauge can be made to order, specify circular (up to 6 elements for carrying in a bucket) or linear (with its own housing and optional empty position for post filter)

Bulk head connector, suited for mounting a tube to the bottom of a bucket 1/4" hose barb $4, with holder for prefilter or screen $9 WR3/8" hose barb $5, with holder for prefilter or screen $9 WR1/2" hose barb $4.50, with holder for prefilter or screen $10 WR

Life gauge - a flexible tape measure (Katadyn suggests the end of life is when the diameter is 1.5" but if there are no thin spots, a low-pressure system can use them a little longer. If you find any micro cracks in the filter element, you reached end of life.) $1 WR

Cleaning pan, don't forget to have something to clean the Katadyn elements in. It is best to lay the filters horizontally partially submerged when cleaning but a 5 gallon bucket would do.

Source Codes

McM McMaster-Carr Supply Co

POB 4355

Chicago, IL 60680 708-833-0300

CP Cole-Parmer Instrument Co

625 E. Bunker Court

Vernon Hills IL 60061 800-323-4340

Leh Lehman Hardware

POB 321

Kidron, OH 44636 330-857-5757

RC Rockland Corp

12320 E. Skelly Drive

Tulsa, OK 74128 800-258-5028

WR Wayne Robey

215 S. 6th St

Lafayette, IN 47901 765-742-7850

All prices subject to change without notice, prices marked WR are only intended for those participating in Rainbow gatherings since this is a not for profit undertaking on my part. Others may inquire. Shipping is additional unless otherwise indicated. I can supply any parts listed herein but for additional shipping charges which will vary with my inventory.

Revised 03/10/98 by Wayne Robey

Edited and converted to MS Word 7/9/98 by Zhahai Stewart