Hyperflow kit, courtesy MSR.
To be legally called a 'purifier,' a unit has to meet the EPA specifications for blocking the passage of protozoa (large), bacteria (medium), and viruses (small). The industry term 'microfilter' does not have a legal definition, but is generally taken to mean a filter which meets the EPA specifications for bacteria and protozoa, but not the specifications for viruses. In general, viruses are just too small for small filters.
Most microfilters use either a labyrinth filter (a maze of twisty little passages, all the same) or an absolute filter layer (holes no larger than the filter rating). The latter may be referred to as a membrane. Both sorts have the disadvantage that they block up fairly quickly when handling murky water, and you have to either mechanically scrape the filter cartridge surface clean or replace the cartridge.
End view of the filter cartridge showing the tubes.
This MSR Hyperflow filter is a little different. Instead of having a single cylindrical membrane, it has a mass of fine 'micro-porous' tubes all bundled together. The tubes have tiny 0.2 micron holes in the walls for the water to flow through: holes small enough to block the bugs. Water is pumped from the outside of the tubes into them, to emerge at the open end. MSR calls this a 'Hollow Fiber Membrane'.
The photo above shows the ends of the little tubes. They look as though they are packed very loosely in the block of epoxy at this end, but this is a bit deceptive. The tubes do not go down to the other end of the cartridge and stop there: instead they are actually folded over at the bottom so that both ends are at the top face shown here. If you peer inside the other end of the filter cartridge, you can just see the fold of the tubes.
The filter cartridge seems fairly robust, although MSR cautions against dropping the cartridge by itself onto hard surfaces. If it is dropped, it can be tested by the user: instructions are included for this. MSR also cautions against letting the wet cartridge freeze. Ice is likely to rupture the fine tubes, so this unit is not suitable for winter trips.
So far, this filter unit has strong similarities to other membrane filter cartridge units. But this MSR Hyperflow cartridge can be 'back-flushed' with clean water. The idea is that back-flushing causes most of the stuff sitting on the surface of the tubes - dirt, algae and bugs, to be dislodged from the surface. This should leave a clean surface and restore the filter to nearly full capacity. You can also disinfect the filter with a very dilute solution of household bleach for storage.
For those curious about the details of the mechanism, it seems that when the filter is back-flushed, the pressure inside the tubes increases, which forces them open just enough that particles wedged in the very small holes should fall back out of them and go back out the inlet port of the filter. Of course, you must use clean (filtered) water to do this!
The filter package consists of the filter itself, an outlet adapter for the standard Nalge bottle thread, a long silicone rubber hose and a float to go on the end of the hose. There's the inevitable stuff sack, of course, with a sewn-in set of instructions about how to arrange the valves. Finally there's a wrapped bundle of instruction leaflets in a huge range of languages.
MSR claims that the Hyperflow weighs 209 grams (7.4 ounces), but the full kit is much heavier than that: over 300 grams (10.6 ounces). I was able to get a 'minimal' kit down to about 220 grams (7.8 ounces), although the exact weight varies depending on how much water is left inside.
The underneath of the inlet filter.
The inlet filter is a neat idea. Instead of a small bobble or acorn which floats underwater, this filter floats face down on the top of the water. This gives it a largish surface area and also lets you get water from very shallow sources. Rub the filter face into the mud, and it will block up, of course, but it's actually a fairly tough nylon mesh which can be easily back-flushed to clear muck off the surface.
The silicone rubber tube can be detached from the inlet filter, so the latter can be replaced. The hook and loop tapes attached to the inlet filter are meant to allow you to bundle the whole thing up neatly, but they seemed to be more of a hindrance than anything else, really. They thread through slots and can be disconnected very easily, too. After a very small amount of use, the tapes were removed from the field test package and the weight quoted below is without them.
The Nalge-compatible cap or connector looks neat, but it has several deficiencies. For a start, at 37 grams (1.3 ounces) it seems a bit like excess weight. In addition, having the filter connected this way to the bottle turns out to be very inconvenient during filtering for a couple of reasons, which will be explained later. Finally, the connection between the pump and the cap is not watertight, so when you tip the Nalge bottle upside for back-flushing, water dribbles out everywhere.
Katadyn Hiker outlet hose and hook.
One alternative to the cap idea is use a second bit of silicone hose to go from the outlet to whatever water bottle you are using. There is a far more useful spigot in the middle of the outlet fitting which will conveniently take such a silicone rubber hose. If you don't have some spare silicone tubing handy, you could even cut off about 200 - 300 millimeters (8 - 12 inches) of the long inlet hose and use that. You would need to keep this outlet hose clean, of course: it could, for instance, be carried in a very small ziplock bag. It would be nice if MSR could also provide a 'hook' to make the hose hang on the neck of any bottle, like the one which comes with the Katadyn Hiker, shown above.
MSR claims that the outlet fitting will fit into 'most water bottles'. It may fit in some, but it is just too big to fit into the neck of the 1.25 liter fizzy water PET bottles I use all the time. However, I don't think this matters, as the whole idea of a direct connection turns out to have relatively little practical value.
The pumping action.
Basically, you put the inlet filter in the water source and pump. The filter has a pump action which - when new, is a bit softer than some other models we have tested over the years. The action needs to be soft if it is to be pumped between one's arms as shown here, but you can hold it in other ways of course. The flow rate is quite good - when new, for all the low force involved, and this is due to the rather large surface area of the filter - all those little tubes of course. MSR claims one liter every 20 strokes, or 'more than 3 liters per minute'. My maths says that means they expect you to do 60 pumps per minute, or one pump per second. Hum - that might be a slight exaggeration of my arm strength, especially at the end of a hard day. More on this later.
Sooner or later the filter will block, and it's then time to strip it down for back-flushing. What this means is that you have to reverse two small valves, one on each end of the pump. One is at the middle of the filter and one is inside the inlet housing. You unscrew the lower part of the pump to reveal one of the valves, as shown below. You take the valve out, reverse it, and pop it back in. Then you repeat this at the inlet end, inside the black lump. That valve is slightly larger and can be extracted using two long bits sticking out from the face of the valve. The instructions are clear about how to do this.
The valve in the middle of the filter.
Then you feed clean water into the outlet end and pump to 'back-flush' a bit. The accumulated muck should then fall out. The main instruction booklet covers the back-flushing process quite well. In addition, there is a second instruction leaflet which includes detailed instructions on an easily-done test the user can perform to confirm that the filter remains intact. This attention to easy field maintenance and user confidence is gratifying.
Formal Laboratory Tests Results
We requested information from MSR about laboratory test reports to demonstrate compliance with the EPA specifications. The MSR reply was as follows. We feel that MSR's reputation is good enough for this to be an acceptable assurance at this stage.
"As for the test reports, I don't have anything available to give you at this time. We are in compliance with the EPA Guide Standard for the removal of bacteria and protozoa. We have tested the filters rigorously in-house, meeting a variety of internal performance specifications as well as the NSF P231 standard for the removal of bacteria and protozoa. We will have test reports available on the website soon, but I do not have a timeframe. For this I apologize, but we'll have it available as soon as possible."
Well, of course, the first thing I did was to try to access the valves and the end of the filter. I tried... and failed. I simply could not get the pump to unscrew. This prompted some emails with MSR, who responded very quickly. To cut a long story short, I was testing a manually assembled pre-production unit which had somehow acquired one O-ring too many. The excess O-ring was blocking the mechanism needed for unscrewing the pump - but I didn't know that at first.
MSR sent a replacement unit without question. Once I had pulled that one apart I could see where the problem was, and was able to disassemble the first unit and extract the errant O-ring. I sent a photo to MSR, and they were quite surprised. But they were prompt to reply to all my requests for assistance. So Backpacking Light ended up with two functioning units for testing. This allowed the testing to be done by two separate Editors.
Filtered and unfiltered water.
The first brief test was conducted using water from the dam on my farm. The dam has ducks swimming in it, and sometimes looks a bit murky. Typically I can see a few feet down through the water, but not much more. We use this water for all our family needs, except drinking water, at home. It is pumped up from the dam through a neat Israeli filter which we clean out once every couple of months. There is some suspended algae, but not a huge amount.
You might say that I should be testing the unit with better water than that, but this is not so. If I am high in the mountains and collecting water from a little stream, I usually don't bother to filter it. It is when I have to get water from a source like the dam that I want to use the filter. (We use this water for everything in my house except drinking: we have a rainwater tank for that.)
The filtered water is on the right, while the unfiltered water is on the left. The plastic in the front is a shiny white. The filter took out all the fine algae that was in the water (which the photo does not really show), but obviously did not take out the tannin color. That's okay: tannin doesn't hurt. The pumping action was not difficult over this volume.
This test was done using the Nalge bottle connector, into a two-liter rectangular Nalge bottle. It was not the easiest way to use the pump: I had to drive it between my arms, with the bottle hanging off one end. It was clumsy to do and hard work after a while. From experience with many other filters, I know it is much easier to pump downwards, with one end of the pump on the ground. That method lets me use my shoulder muscles to drive the pump, and it is much easier with this pump, too.
However, pumping vertically means I can't use the Nalge connector. Instead, I used a length of silicone rubber tubing to go from a spigot at the outlet into the bottle. The hose is visible in the next photo. I found when doing this that the hose MSR uses on this filter is actually slightly larger than that used on some other filters, and that smaller hoses will slip off the spigot fairly easily. But a bit of the hose supplied by MSR sticks on fairly well if pushed on firmly.
Having an outlet hose in place means that the filter can be used vertically. If you try to use the filter with the inlet at the top, you will never get rid of all the air in the pump. It will remain there as an energy-sapping 'air-spring'. Always have the inlet at the bottom: the air will be at the top and will get quickly flushed out in a couple of strokes.
What the vendors never mention about routine operation of a filter is that the cycling pressure and flow in the inlet tube can make the pre-filter dance across the water. It happens to some degree with this filter too, with the added complication that the pre-filter can easily flip upside down. What I usually do when using such a filter pump is to lock the inlet hose down somehow: with a rock or my foot. The pre-filter still dances, but it stays in one place. Just don't hold the tube too hard or you will squash it.
The filter set up for back-flushing.
Then I tried the back-flushing process. The bottom valve, in the rounded inlet housing, is tricky to reverse. It has to be put back in very square to get the valve bit into the recess provided. The other valve, at the middle, is easy to reverse. Both get easier with practice: it is not a problem.
MSR shows a diagram of the filter with a Nalge bottle attached and inverted for the back-flushing process, but they fail to mention that the connection between the outlet spigot and the Nalge connector is not sealed. As mentioned above, it can leak all over you. I used the outlet hose I had rigged up to feed clean water from the Nalge bottle to the filter outlet, as shown in the photo above.
When starting to back-flush, there is usually air in the system, and the little filter tubes don't like passing air. It is hard to flush the air out of the pump, and MSR recommends that you flush any air out the inlet tubing first. I found that the best way to hold the filter during back-flushing is as shown in the photo above: 'upside down'. Doing it this way puts any air in the system at the top, where it gets flushed out quickly.
Even so, sucking water back through the filter requires that you pull upwards on the pump section (the top red bit), and this is much harder than leaning downwards on the pump to push down. If you try to do this too quickly you can get air leaks from the middle O-ring seal as well. If I pumped very fast I could get quite a vacuum in the pump chamber (half the volume in fact), and that seems a waste of energy. It is better to pump slowly. After a few slow strokes of the pump I managed to clear the air out, and by taking it slowly I was able to run the recommended half a liter of water back through the filter. It isn't really difficult, just slow.
I was able to clean the flat red pre-filter very easily. I took the hose off the inlet and allowed the water in it to drain back through the pre-filter. Then I blew gently down the hose, making the pre-filter fabric bulge slightly. That seemed to dislodge just about everything. This is something I would possibly do every time I used the filter - at least it gets rid of the water in the hose.
The middle O-ring gets a lot of action, and needs to be lubricated regularly. For some strange reason, MSR does not include any silicone grease with the filter kit, although it is available in the recommended (and very comprehensive) maintenance kit.
We report here, three different rounds of field testing: in the field with clear water by the two authors and a more controlled test using dam water.
Field testing, with clear water and the Nalge bottle connector.
The photo here shows the Hyperflow being used with the Nalge bottle connector on a shallow clear mountain steam. There is some (harmless) iron bacteria on the bottom of the creek, but the floating inlet filter handled that excellently. You can see the inlet tube being restrained by the author's foot: this kept it in place and stopped it from flipping. You can also see the awkward position of the author's arms dictated by the use of the Nalge bottle connector.
The filter worked satisfactorily here over a couple of liters, but I would note that I probably would not normally bother filtering this clear mountain water. The inlet filter did pick up some of the red iron bacteria stuff from the creek bed, but this blew off easily when I back-flushed it.
Will Rietveld and his wife used the MSR Hyperflow water filter on three backpacking trips totaling eleven days. In his notes, he wrote:
"On each trip we solely relied on the Hyperflow, and all of the water we filtered was from clear mountain streams. First of all, it weighs a bit more than specified; we carried it in its zippered carry bag, and the complete kit (including a cap for a Nalge-type bottle) weighs 10.1 ounces, which is a bit more than the specified 7.4 ounces. So much for light weight!
From the beginning, the filter was fairly stiff to pump, and it has stayed that way. We back-flushed it after every trip, but back-flushing made no noticeable difference in its operation. Because of its stiff pumping and the need to pause briefly to allow the pump chamber to fill up, the pumping rate we achieved was about one liter per minute, much less than the three liters per minute claimed. The pump chamber filled up faster if we held it lower to the ground, but that made the pumping even more difficult.
Filtering directly into a narrow mouth water bottle required some downward pressure on the delivery end of the filter to keep it in position, which was a little awkward. The special cap provided that fits a Nalge-type bottle or flask works beautifully; we used it on a six-liter Nalge Canteen, then filled other bottles from the larger container. There is a fitting on the top of the cap that the outlet of the filter locks into, which helped to streamline the filtering process, especially for one-person operation.
Overall, we found the MSR Hyperflow water filter to be reliable, but laborious. At 10.1 ounces for the whole kit, it's no lightweight compared to other filters, and it seems heavy and complex compared to chemical water treatment. Our main reservations about it are the higher force than expected to pump it, and the relatively slow delivery rate. For us, it was a lot less than hyperflow."
Thus both authors found that the Hyperflow worked reasonably well with clear mountain water, albeit with more force and more slowly than claimed.
Finally, a controlled test was done with my dam water, as described above. It should be remembered that the dam water is not crystal clear, but it is the sort of water you would want to filter if you had to use it. The cycles were as follows:
Filter 3 liters.
Pumping action started to get harder after 2 liters.
Back-flush with five good strokes.
The first two strokes were difficult, then it got easier.
Filter 3 liters.
Again, pumping got harder after 2 liters.
Back-flush with five good strokes.
The first three strokes were difficult, then it got easier.
Filter 2 liters.
These were harder to filter, and pumping at any speed collapsed the inlet valve several times.
Back-flush with ten good strokes.
These were hard to pump.
Filter 2 liters.
These were quite hard to pump, and the inlet valve collapsed several times.
While back-flushing, I found that it was easily possible to get half the pump chamber effectively holding a vacuum. I had to pull the pump up and then wait while water flowed through the filter to fill the pump chamber up. This could take a couple of seconds. Clearly the backwards flow was not good.
MSR says to back-flush with 0.5 liters (ten strokes) each time. In this test I only used 0.25 liters (five strokes). However, if I have to use 0.5 liters to clean out the filter after every 3 liters, then the whole thing becomes a serious waste of effort. I question whether there would be a lot of difference between the number of strokes I used and MSR's recommendation anyhow. If five strokes won't clear the filter, I doubt that ten would.
You will see my comment that the inlet valve collapsed during the testing. This means that the soft rubber flaps inverted, and let water go back out the inlet hose. This only happens when the pressure inside the pump chamber gets too high for the valve. I didn't think I was exerting all that much force. This suggests that the design of the valves needs some attention. I see this as being a 'Version 1.0' problem.
Not mentioned above, I also observed that the outlet valve was doing something strange. When I pulled the pump back to refill the pump chamber with more water from the inlet valve, the water would recoil a distance up the outlet hose. I really do not understand what was happening there.
Finally, I tried using a Katadyn Hiker filter on the same water. The Hiker is widely regarded as the benchmark filter against which other filters are often compared. It uses a large pleated filter with a very long life, even on this sort of water. (I would expect to get at least 100 liters through it here, and probably considerably more.) The Katadyn Hiker was easier to pump than the MSR Hyperflow.
The design of the MSR Hyperflow is nice. The filter is light. But the filter element seems to be just not up to the task of handling anything other than clear water. To be sure, some 'clear' water can be problematic, and the Hyperflow does handle it moderately well. But that is not good enough for walkers who often have to handle a wider range of water quality with little in the way of choice.
My understanding is that MSR is very well aware of this problem, is concerned about it, and that there may be an updated version of the Hyperflow in the future. We look forward to seeing the updated version. In the meantime we cannot recommend this filter for use on anything other than clear water.
|various plastics and polymers|
|various plastics and polymers|
|0.2 micron. Removes bacteria and protozoa to a degree which meets EPA requirements for water purification. Note that this does not include viruses.|
|~220 g (7.8 oz) for minimal kit of pump, full-length hose and inlet filter|
|Filter Kit: US$99.95
Replacement cartridge kit: US$39.95
Replacement prefilter: US$14.95
Maintenance kit: US$19.95
- A fairly low weight
- Neat design
- Can be repeatedly cleaned easily - up to a point
- Useful inlet filter
- Cartridge and inlet filters can be replaced
- Removes bacteria and protozoa to EPA specifications
What's Not So Good
- Quickly blocked with water which is not really clear
- Valve design collapses under pressure
- Nalge bottle connector is not ergonomic - and leaks
- Needs an outlet hose and hook added to the kit
- Unsafe for winter use when freezing is possible
- Silicone grease for the O-ring is not included with the kit
- Does not remove viruses (most microfilters have this limitation)
Editorial Addendum (02/06/09):
Note from the manufacturer:
We have identified a flow performance issue with some of the hollow fiber filter cartridges contained in MSR HyperFlow microfilters. The performance issue DOES NOT affect the product's ability to filter safe drinking water, but can be frustrating, as the flow rate of the filter may not perform to product specifications. The issue has been rectified, and all filter cartridges currently in production for the MSR HyperFlow microfilter perform to flow specifications. We have worked with our retailers to replace units they have in stock that may have this issue. Any consumer that is experiencing less-than-expected flow rates on this product right out of the box or after back flushing is urged to call us at 1.800.531.9531, and we will send a valid replacement filter cartridge at no charge if the original was manufactured prior to November 11, 2008. (Please have the filter element handy, as we will ask for the serial number for our records and manufacture date verification.)