DIY Gavity feed water filtering

[SkeeterPee]

Looking for some input.

I have a sawyer mini, might consider getting a squeeze since so many say it is better, but so far the mini has worked for me.

I have a 32 oz sawyer bag which makes filtering easier. I don't find squeezing the bag all that hard, but it is cold on the hands later fall so I want to set this up as a gravity system.

I have watch some videos, but I have not seen anyone say what works best. I assume there are some engineering principles involved that affect how this works.

so some questions

I plan to filter with a small carboy filter too right after the sawyer unless there is a better place for the carbon filter. Currently that tubing is about 2 inches between filters.

does it make sense to have the filters connected directly to the dirty water bag? or is there some length of tubing that should between dirty water bag and filter.

Likewise should there be a length of tubing after the filters before going into a smart bottle?

I am thinking a full tube of water after the filter should help pull more water through the filter. is this true? or does work better to have the long tubing between the dirty water bag and the filter? how much length of tubing before or after filter?

sure I can experiment, but I bet some you men and women are smart enough to know the answer!

Thanks, pat

[HooKooDooKu]

What I did was to get an Evernew Water Carry for dirty water, and a Platypus Hoser for the clean water. Simply remove the bitevalve from the Hoser, and you can then push the end of the hose over the output nipple of the Sawyer Mini. However, the hose that comes with the Hoser is pretty stiff. So instead, find some of the stretchy medical grade hose (in my case I simply reused the hoses from my old SweetWater Pump system). To replace the stiff hose on the Hoser, you'll likely have to carefully cut the hose where it attached to the barbed nipple. Simply slide the new hose over the barb from the Hoser. Ok, it's not going to be THAT simple. It's a pretty tight fit. But a trick I learned from an Irrigation forum for putting hoses on barb fittings is to get some water base lube from the "intimacy" isle (think KY). The lube makes it easy to get the hose on, and by using a water based lube it is easy to clean the lube off.
Here's a picture of my test setup in the kitchen (http://whiteblaze.net/forum/vbg/showimage.php?i=60632&catid=member&imageuser=11871) (used a hole punch in the corner of the dirty water bag and some thin rope)

[Roamin]

Pat, I am also making a gravity system. There are a lot of videos on YouTube. I am taking a little from each video to make my set up. May I ask why the two filters?

The one bit of advice I like is to not have your clean bag directly below your dirty bag in case of any drips. Having a slight bend will allow water to drip off the tubing before getting to your clean bag. I am not too concerned with most of the water on the AT, but it is still a valid point in my opinion.

[SkeeterPee]

there are lots of videos, But none go into the science of what works best. I expect some designs work better than others. is it better to have long hose before the filter. or after the filter or on each side. I think it matters with flow rate. the weight of a column of water in the tubing would either push water through the filter or pull it out if after the filter. certainly you can see if you attach filter directly to your bag it does not work well. So that column of water is important.

The second filter is a small gravity works carbon filter which would filter some organics out of the water for a better taste. Though the water I have been getting so far has been very good so I probably wouldn't really need that. They are good for only about 300 gals.

[nsherry61]

Your head of water needs to push through the filter, not pull (long tube above the filter). The more head, the faster the flow. And, you'll probably give up on the second filter unless the water is really bad.

[Five Tango]

If you want to filter twice you might like to consider one of these.I have one in line with the Sawyer squeeze.The longer the hose the higher the dirty water bag and more total dynamic head for better flow.Note the Red Line filter is rated for 120 gallons.I wonder if prefiltering with the Sawyer extends that life expectancy or not..........
http://www.campsaver.com/frontier-series-ii-red-line-replacement-filter

[SkeeterPee]

Thanks for the tips.

[Mudsock]

The physics behind the two bag gravity filter setup is pretty simple. The length of the vertical column of water between the entrance of the clean water bag at the bottom and the top of the water in the dirty water bag determines the pressure of the water at the clean bag entry. The pressure with no flow is simply Rho g h, where Rho is the fluid density, g is the gravitational constant and h is the height. So, neglecting the height of the water in the dirty water bag, static pressure is directly related to the vertical length of the hose between the two bags. The filters have no impact on pressure WHEN THERE IS NO FLOW. When flow occurs, there will be a pressure drop across the filter(s) directly related to the flow rate. There is also a pressure drop linearly along the tubing due to friction on the walls. Of course, with no flow, there is no friction. The "circuit" can be envisioned at any fixed flow rate by viewing the pressure as a battery, producing a Voltage, the filter(s) as a resistor, the hose as a distributed resistor along its length (much like wire resistance) and the flow of water as current. Filter resistance will be the larger resistance with reasonable sized hose. The total of all resistance can be summed or lumped, regardless of where it is in the series circuit and be represented as one resistance or resistor total. Ohm's law is a simple way to look at the circuit. Voltage = Current x Resistance, or Pressure = Flow x filter and hose resistance. Another thing to remember is that with increases in flow, resistance changes. That is true in an electrical circuit also, as the resistor heats, it changes resistance. With some materials, it is dramatic and non-linear.

This is a simple plumbing problem with a useful electrical analogy. Assuming the filter and hose have no air voids so that the column is continuous water, the position of the filter makes no difference. It represents a pressure drop based upon flow rate. Consider sliding the resistor along the length of wire connected between the two terminals of a battery. The position of the resistor makes no difference on the current.

The pressure will drop as the dirty bag empties because the height of the water in the bag is part of the height of the column. The decrease in pressure as the bag empties will be proportional to the height of the bag and the vertical length of the hose. Something else to consider with this problem is how much pressure is applied by compression of the clean water bag? I tried filling my Platypus Hoser reservoir while in my Arc Blast. Compression from the tightly filled pack caused an extremely long fill time and a less than full bladder. Opening the pack and removing items until there was no significant pressure on the reservoir made a huge difference.

An interesting part of this problem is that as tubing is added for increased vertical height, the resistance with flow continues to increase. At some point, adding more tubing will increase resistance at A GIVEN FLOW RATE so that longer tubing will offer diminishing increase in pressure. With each diameter of tubing, the maximum effective length will be different. It works like garden hose, or a fire hose. For a long run, they use larger hose because of the pressure drop due to friction. Remember that when you snap open the nozzle with your garden hose, the water shoots a long distance, then quickly drops back to the steady flow distance. That is because of the dynamic pressure drop due to friction along the length of the hose. The drop in pressure is controlled by diameter. If you want a higher flow rate on your bag system, use larger diameter hose. There will quickly be a limit to the benefit. The municipal water system has larger pipes closer to the water supply, as flow rate is highest there. In our neighborhoods, a much smaller pipe serves well. Going from the pipe to the house, a one inch line is enough. The driving factors in all of this is pressure, flow and friction.

There shouldn't be much more to designing a gravity filter system. Limit the number of fittings for highest flow rate and quickest fill times. Each fitting will provide an increase in series resistance and a pressure drop, providing the minimum diameter is lower than the diameter of the hose. Fittings generally are pressed inside of the hoses, so they always reduce diameter. We could work around that problem with rigid tubing. That would be difficult to carry on the trail.

[HooKooDooKu]

Shouldn't matter if your pulling or pushing, the water head will get the water thru the filter. If you're using a Sawyer, you want to attach the dirty water bag right to the filter for the simplest design. Of course that means you won't have built you your water head until the tubing below the Sawyer fills up. With a 3' tube, you should have about 1psi pulling the water thru the filter.

[Slosteppin]

After reading the post from mudsock I better understand why my system works.

I use a 4 L dirty water bladder from an old Platypus gravity system. After playing with the Sawyer mini and tubing I had I bought 4 sets of quick connects from REI and more tubing. I have 3 or 4 feet of tubing between dirty water bladder and the filter. Then another 2 or 3 feet of tubing down to the bladders I use for clean water. I usually fill 2 one L bladders and one 2 L bladder. I generally use the same tubing from the filter to clean water bladder as I use for drinking while hiking.

I did a lot of experimenting first in the kitchen then, after mopping several spills, in the backyard, before I got a system that works with no leaks or drips.

[Casual]

I think you're overthinking this. My setup works like this:

Evernew 1500mL "dirty" bag -> Screws to Sawyer Mini -> permanently attached 6" or so of hose, with a male Camelbak quick connect -> female quick connect on Camelbak bladder (bit valve snapped off).

Lay your pack on the ground with the bladder still in it, hang the evernew from a convenient branch or whatever. 2-3' up is plenty to drain the bag in <5 mins.

34075

[Mudsock]

Quote: Assuming the filter and hose have no air voids so that the column is continuous water, the position of the filter makes no difference.

This is where the analysis VERY often deviates from our experience. If we let the water run out the bottom of the tube just at the clean water bag long enough to get the air in the filter and tubing flushed and then put the tube on the clean bag, we have that continuous column and a much higher flow rate. But with a highly restricted filter, the column does not remain continuous water if there is air in the lower clean bag. A way to prevent air in the bag from getting in the tube would be to form a trap with the hose, like under the kitchen sink. This is practical for a stationary system; not easy on the trail. Air in a hydraulic system has to be kept out of the lines.

If the pressure drop across the filter becomes enough that the flow decreases to the point where air from a partially filled bladder could get into the tube, it would effectively remove most of the suction from the bottom end of the filter. Water is in-compressible and does not change volume unless it changes state, i.e. turns to gas as in cavitation - a vapor void. That won't happen to us. So, as long as the column is water, the model holds. If the fluid in the column at any point becomes air, Rho goes down drastically at that point and for the entire length of the column of air. Water will trickle through air. We now have distributed batteries in the model with potentially much less total voltage, i.e. pressure. Air is highly compressible and we cannot maintain the same dynamic as with water. When air compresses, Rho (density) changes again. The suction from below becomes highly elastic or spongy, like when brake lines have air in them. We need other circuit elements in that case. Simple resistors are not enough. At this point, I think we need to add a capacitor in parallel with the greatly higher valued resistor that represents the air. It is no longer a simple steady state DC problem, but rather a DC flow with a smaller magnitude AC component superimposed. The sign of the current never changes, yet it undulates.

When air is present in a portion of the tube, things change. Ideally, the air is below the filter. If that is the case, then having the filter at the bottom will keep the greatest pressure drop across the filter and maximize the flow. It would seem that if having the filter at the bottom makes a difference, then there is always air in the column.

I guess I didn't get all the factors included. To simplistic.

Testing at the kitchen sink is the way Edison would have done it. He often failed, but got it right in the end.

[RangerZ]

I was going to write the Bernoulli equations for the high and low filter cases but I don't think that I will now. Be careful of any back pressure from a compressed bladder.

I put the filter right below the dirty water bag, not the best place hydraulically but the easiest. I've drilled a soda bottle cap to fit over the outlet nipple on the filter to allow me to filter right into a Sawyer bag. I also have a quick disconnect on a short piece of tube on the outlet nipple that I use to filter into the drinking tube for the bladder in my pack (much like pictured above).

[SkeeterPee]

Thanks for all the information!

Based on above I was thinking of getting 10 ft of tubing which is probably overkill and would end up using less. I was checking home depot for 1/4 ID 5/16 OD and did not see it. will stop inside to see if I missed it. any one have a good link? I have tubing for my wine making which might be 1/4 inch, but I since I use vacuum racking, I expect is thicker walled than needed.

Also if I understood what was said about the air pockets are that they are very bad. And it might be easier to have a longish tube above the filter since it is easier to keep that full with no air bubbles. Then it seems like it might be harder to keep a column of water after the filter, but if you can, then it is equal to water above as far as flow rate.

I have been using water bottles for storing clean water. 1-2 liters total. I have thought of using a bladder again. Would it be better to filter into a bottle like a smart water bottle, and figure out a cap connection that can be used to squeeze that into the bladder so you don't have to remove from pack? the back pressure of filtering into a bladder in the pack would probably be a problem.

[HooKooDooKu]

Unless you plan to climb a tree to hang your dirty water bag, 10' of hose is WAY OVERKILL! Something on the order of 2 to 3 feet is all you need. I can see various setups where 5 to 6 might come in handy. But 10 is just adding weight to the system... will likely out weight the filter.

[SkeeterPee]

10 ft would be so I could experiment. I was not planning to use that much.

[capehiker]

I use a DIY gravity system that cost me $3.00. I don't really think about the science aspect of it (but it is pretty cool to learn), I just want simple, easy, and light when on the trail. My DIY system filters fast enough to where I've never considered buying a full gravity based system. It works great at rest stops and camp at night.

Here's a quick video of my DIY gravity hack:


https://www.youtube.com/watch?v=yk62Qe6buE0

[nsherry61]

I love seeing science and engineering theory enter into the real world! Great electrical analogies guys.
1) Yes, it is a hassle to get the tube completely filled with water after the filter to start gravity filtering, and, it is incredibly difficult to avoid air in the tube after the filter, and thus "pulling" water through the filter with a long hose after the filter tends to be quite problematic in practice, even though it should work in theory.
2) A water column pulling through a filter (or other restriction) will "break" with about 6 feet of height, so it is truly impossible, even in theory to "pull" more than a six foot column of water (the reason well pumps are submerged in the well instead of sitting above ground).

So yes, put the dirty water reservoir as high above the filter as possible to maximum flow rate.

In practice, it seems like most people start seeing diminishing returns with more than 3 or 4 feet of tubing, maybe because the flow rate through the filter is not linear relative to the amount of pressure, so doubling the pressure has much less effect on flow rate through the filter going from 2 to 4 feet of head than it does going from 4 to 8 feet of head?

[kenl]

My experience is the a Platypus gravity system. (To ensure I've got this right, you can visit their website for user manuals as well).

The two liter gravity system is as follows:
- dirty bag with quick connect
- about eight inches of hose. One end has a quick connect for the dirty bag and the other is attached directly to the filter
- about three feet of hose. This provides sufficient gravity feed (as noted by other contributors) and also allow you to raise clean bag above dirty bag if you need to purge air bubbles from system.
- clean bag (or smart water bottles). In my case, I put a quick connect at the bottom of the hose coming from the filter and hooked this up directly to the quick connect on my hydration bladder. The quick connect also fits into a smart water bottle so at night I can fill up bladder and bottles to ensure sufficient water for dinner and breakfast.

Hope this helps. Safe travels.

[SWODaddy]

I've found that the Platypus "Dirty Bag" works best for me.

34087

The large "zip" opening means that I can collect water from areas with low flow no problem. But the seal is still strong enough for me to roll the top and add some pressure if I'm in a hurry. It uses a generic female hose adapter - I bought some camelback fittings from REI which work fine with it. Platypus makes a kit which accomplishes the same thing, but it's more expensive and I'm not sure the filter will screw onto a bottle like the sawyer mini will.

[poolskaterx]

I use a Sawyer screwed onto the platypus that is used as the dirty bag, hang it off a tree branch or hiking pole, plug in my bite valve of my drinking system (2 one liter platypus bottles with drink tube) right into the nipple of the filter and give the dirty bag a little squeeze to get it started. Viola!
Dirty bag $12
Drink tube with bite valve $10
2 platypus 1l bottles $24
Sawyer filter $21

total system less than $70, engineering degree optional:)

[rickb]

Just a few photos. Number 3 was from my home brew store, and number 4 were from REI store. What is most interesting to me is that while the length of the hose matters, the position of the filter relative to the input bag does not.

34229342303423134232