Specific DE filter questions PLEASE....

[SB Brewer]

We have used a 3m sized DE filter from spaldini for several years. The problem has been the same since day one and we have tried many things and ratios of DE and dosing along with trying hi flow to other flow levels of DE. The problem of which I speak is that I only filter about 2 or 3 barrels of beer before the filter sends cloudy beer again. So we try to recirculate and add some dosing but it does not help so we emtpy and clean up and start over. For 10 bbls of beer, we set the bed and clean the entire thing 3 times. This is an 8 hour filter of only 10 bbls of beer but the manual from Spaldini says it can do more than 10 bbls in an hour once the bed is set. So I have specific troubleshooting questions please.

1. When the beer is coming through clear and then after a 2 or 3 bbls the beer starts to be cloudy coming out, AND the PSI on the bell is still only 3 to 5 PSI. Is this lack of DE dosing or too much DE dosing? Or has nothing to do with the amount of DE?
2. How is it that the beer comes out cloudy and then to drain the bell we push the beer out of the 50 gallon bell by co2, it comes out clear from the bell?
3. If there is a pressure build up in the bell during circulation, does that mean not enough DE or too much DE or too much yeast? Ratio issue or not?
4. Does anyone know of a troublshooting guide to DE filtering somewhere on the web? Or videos? I have been searching for years and have read and watched all I could find. But nothing explains what to do when PSI is building up or when Cloudy beer comes out while PSI is not building etc...

Thank you for your time..
Pete
Storybook Brewing

[gitchegumee]

Hey Pete! Filtering with DE requires a very thorough understanding of depth filtration, equipment build, fluid dynamics and kieselguhr/particle size relationships. Hard to know exactly where your problem is without being there to see what's going on. My guesses to your questions: 1) nothing to do with amount of dosing. 2) Could be many reasons. 3) Usually too much yeast & insufficient clearing of fermenter bottoms. 4) Nothing will tell you what to do for every tiny thing that can go wrong.

So, if you break through your precoat after a small increase in differential pressure, my first reaction is to look for tiny holes in the septum. You should be emptying your filter very carefully and then opening the bell to examine the leaves thoroughly. Look for dimples in what should be a perfectly flat precoat surface. I've patched many tiny holes formed by a broken wire. Don't disassemble the leaf stack unnecessarily. They are very delicate and easy to damage. Incorrect reassembly can easily cause small leaks. Don't know if you have an erosion-style unit, but throttling the discharge of the dosing tank is wrong. You should only throttle the inlet to the dosing tank with full open outlet. I've seen valve seals scored badly from the abrasive nature of kieselguhr. Check that your machine is in perfect working order. I'm guessing that it's the machine, not the technique. Best of luck!

[SB Brewer]

Thank you very much. I shall check everywhere.

[13356690973]

Hi mate, we think this DE fileter has some problem so can not bear so much litres of beer filtering process , It has nothing to do with your filtering technology and another reason maybe there is too much yeast in the bottom cone of your fermentation tank........

[SB Brewer]

Thank you very much. I shall inspect all over.

[dick murton]

I am assuming this is a horizontal leaf filter. Whether vertical plate and frame, vertical leaf, or horizontal leaf, many of the conditions required are the same.
You state the pressure on the bell is only 3 to 5 psi. If this is the case, then you are almost certainly getting gas breakout, particularly when the beer comes into contact with the KG bed. The gas bubbles then create holes or even completely dislodge the powder from the bed, so allowing unfiltered beer to pass through the now destroyed bed into the discharge stream.
The operating pressure is normally recommended as 1 barat the discharge, but to some extent this depends on the CO2 content of the beer and the temperature at which you are filtering – check the gas solubility / pressure charts – all around on the internet if you don’t already have one.
Flow rate for both pre-coating and main body filtering. For vertical leaf and plate and frame filters, pre-coating is often carried out at a higher flow rate, up to 150% of beer flow rate, to help consistent precoat layer formation. But this is very much a case of experimentation as I have used different rates for different filters and different powder qualities and grades.
Vertical plate & frame (and presumably vertical leaf) filter - main beer filtration flow rate typically 3.25 hl/m2
Horizontal leaf filter - main beer filtration flow rate typically 6.5 hl/m2

• System pressure 1.0 bar
• Total pre-coat 800 – 1000 g / m²
• Pre-coat mix ratio 1:10 (not less than 1:8 to avoid overloading the pre-coat pump)
• Two stage pre-coat
• 1^st coarse pre-coat 400 – 500 g / m²
• 2^nd fine pre-coat 400 – 500 g / m²
• 2^nd pre-coat normally same mix as dosing
• Do NOT allow pre-coat tank to empty or draw air into the filter system
• Close vents before shutting off pre-coat dosing pump
• Flush through with fresh, ideally chilled water and ensure all air is removed
• Allow to circulate at the same flow rate for 5 to 10 minutes to stabilise bed before introducing beer

By all means PM me with any questions and I will do my best to answer. I have a load of material, but troubleshooting is something that as you have found, is difficult to write up a comprehensive guide.

[SB Brewer]

Thank you very much for the information. This is appreciated. You are using 2 pre coats, 1st coarse and 2nd fine. But on the many sites including this one, it say "DE comes in several grades. Choose it based on the type of solid being removed from the beer. Don’t blend different grades to try to average the particle size. Because DE works by forming bridges, it relies on a consistent particle size to work effectively." So I am not sure I understand. I guess maybe you are saying this is fine because we are not mixing them together. Only layering them? I use Hi Flow to pre coat and dosing, so I assume Hi Flow is Course and therefore, I should get results using only Course? Thanks again.

[dick murton]

The purpose of a coarse filter powder is to bridge the gaps between the individual steel wires and stop finer grades passing straight through the gaps. So this layer is normally about 2 mm thick, possibly thicker. It also means during spin/wash-off, the filter bed is easy to remove as it sits on top of the filter mesh, and is not embedded in it. No problem if it is thicker, but if too thin some of the finer powder, and thus also beer fines will pass through. The second layer is the finer powder used for bodyfeed, and is there to ensure in the initial stages of beer filtration, none of the yeast / protein haze passes through the coarse powder and into the bright tank. If you mix powders, all that happens is that the small particles fill the holes between the fine particles, making an even finer bed, but unfortunately, the degree of restriction doesn't seem to be predictable. Not sure why, but it certainly is not over reliable, though I have worked at breweries where they have mixed powders and got away with it, and others where it has proven a disaster. I assume it is due to variable particle sizes in each batch - the particles are never exactly the same size in a given batch, but the overall porosity of each batch is very consistent.

If you think you may have holes in your plates (I don't - I think it is gas breakout due to unsuitable temperature / pressure) then I suggest that having inspected the plates, you precoat the filter with the first precoat only, drain the filter out as if emptying the filter of beer, and inspect each plate for consistent depth of powder, with no pinholes etc. When you are happy with the first coarse powder precoat, then carry out a full precoat using the proven first precoat procedure, and then add the second precoat. Drain down and check again.

What powder to choose as bodyfeed? In simplistic terms, a coarse bodyfeed powder will not remove as much fine haze material as a fine powder. Underdosing powder, of any grade, will mean the pores between the filter powder particles get blocked more than at correct dose rate, let alone if overdosed, when in effect you are wasting powder and potentially reducing your filter run below that required. The determination of the correct dose rate is trial and error I'm afraid, even in mega breweries people tend to overdose as this generally produces longer filter runs than underdosing.

The quality of the beer you send to the filter determines the dose rate. Filtration temperature ideally minus 1 deg C (approx 30 F), certainly no more than plus 1 deg C (ca 34F). Held at minus one for 24 hours absolute minimum The brighter the beer is before filtration, the longer the run and the better the final clarity. Use of kettle finings and secondary treatment such as isinglass, silica gels (or mixes of both) good turbulent boils, prompt removal of trub, prompt removal of settled yeast at the end of fermentation, suitable mineral ion concentrations in the wort / beer also help - in other words, good brewing practises.

[SB Brewer]

Thank you again. I really appreciate information like this. I did not think about the Mineral ion concentrations in the wort. We use things like Calcium of course but did not connect the protein haze to the filter run length even though it seems obvious now that you pointed it out. LOL Crazy how sometimes I can not see the forest through the trees. Thanks again. I am going to try some things and will post my findings on this chain in case others have similar issues.

[liammckenna]

Sometimes a bit of solka-floc (powdered cellulose) can help plug any small holes in screen. Add at first precoat. Also, you'll want to check the torque on the top retaining nut on the plate pack/screen stack. We were recently having a similar problem with our Cadalpe filter and such turned out to be the problem. It was an easy fix. Good luck. Pax.