What is happening inside the filter? (II)

Main filtration cake from ideal pre-coat

Now that we understand the 3 main possible outcomes of the pre-coat layer, we must understand these are the worst and best possible scenarios, these possibilities are in a line from low flowrate to high flowrate. This means that you can find any deviation in along this line, knowing how the particles behave and how the shape the cake is formed will give you a good understanding of there you are and how close you are to the ideal filtration cake.

Now let analyse what will happen in each scenario of pre-coat layer, when you have another 3 scenarios with the main filtration.

Let start with an ideal pre-coat layer.

The first scenario is, starting the main filtration with an ideal flowrate, it is important to notice that the duration of the main filtration rate can be established in many ways, rather by calculation, by experience or by control instruments. We recommend having a control instruments system where you can measure the differential pressure over filter leaf’s. Then with a control system on the VDS pumps, maintain the right flowrate along the complete filtration cycle.

The first scenario will be a good pre-coat layer with an ideal and constant flowrate, as it happens in the pre-coat layer formation, an ideal flowrate will allow the particles to evenly spread on top of the filter leafs pre-coat layer, with every particle being caught and some point in the pre-coat of the filter cake. Sometimes it is noticed that while the cake grows the DP (differential pressure) increase and the flowrate decrease, therefore to maintain a good flow rate and an even cake in shape and tightness we should control the pump so the flowrate through the entire filtration process is mostly constant. This will allow a nice and smooth drying process and cake drop as well. 

This cake will look like a cake indeed. A nicely even layer on top of the filtration mesh, it will stay on the leafs until vibration forces will allow the filter cake to drop nicely and complete. A cellulose filter aid layer will make sure to protect the filtration mesh and allow the full cake to drop, leaving a clean leaf. Moisture in the cake should be low minimising the product losses on the cycle. 

The scenario where we have and proper pre-coat layer, but run a filtration cycle with a low flowrate, the same scenario of the drop shape cake will happen. Particles settle during the filtration and accumulate on the bottom. Though the filtration quality will be good as the precoat will make sure there is no particles passing by. The shape of the cake will again resemble a drop. In the main filtration, the leaf space should be three times the expected cake(ideally). But when the drop shape happens is possible that the cake will grow so much at the bottom of the filter leaf that leaf to leaf cake won’t be separated from each other anymore.

As you can imagine if this happens or the cake separation space is reduced too much, in the cleaning stage the cake will not drop easily, the cake has no space to move and will stick to the leaf, the cake that drops from the top will also accumulate on with the one already blocked in the bottom bridge creating a big block of stuck cake that has to be removed manually.

Other possible problem, because of the low flowrate, if the solid load is low so the drop shape is not that recognisable or in the case where the particles density makes the easy to transport and don’t settle much, having a low flowrate will create a loose cake. The pressure on the cake is not much and that means that particles are not very tight to each other. Therefore, while drying the cake, the lots of moisture in the available space among the particles, has 2 consequences, first one is the amount of lost product on the cake, and the second one is the chance that the cake is not solid enough to stay on the filter leaf even if you main a positive pressure along the complete filtration cycle.

When this happens, is very easy to notice. When you open the filter, rather than a cake on the leaf you will be dragging mud out of the vessel, damaging the leafs in the process. Regularly this will happen in the filtration of liquid sulphur where the particles are more crystal solids and heavier.

The last scenario with an ideal pre-coat will be when the main filtration flowrate is too high, this result would be in my perspective the worst of the three scenarios we have seen so far considering we have managed an ideal pre-coat layer.  The pre-coat layer is already an efficient filtration media, from the start of the main filtration cycle, it is already catching all the particles no matter the particle size.  

By the contrary of a low rate that may shape a loose cake, the high flow rate will make a too tight cake. This means that particles will be pushed against the pre-coat layer with too much force creating blocked areas fast from the beginning of the filtration. The new cake layer is being created with very low permeability, this decrease the available filtration area, this will increase the flowrate amplifying this sequence. As a result, we will have a filter that will be promptly blocked and won’t allow any more filtration, this will be easily noticed as the differential pressure and the pressure in the vessel will increase rapidly, the pump could just stop working, if the system is not controlled by a flow driven pump this cycle will just be very short.

At this point you have not just finished your filtration cycle too soon but you will also have a hard time drying the cake, sometimes, special when you see this cake forms for the first time, it won’t be easy to notice that there is something wrong, as it would look nicely spread and even a bit moist on the last part of the layer. Most of the times the pump would be blamed for the short cycle.

The reason the cake would look moist on the surface is because the pump trying to accomplish the filtration cycle, will give enough pressure to allow a little longer cycle, but this will be happening at a low flowrate in the cake as its permeability is really low, so the last particles are being pushed with less pressure against each other.

At the end of the filtration cycle you will nicely drop a semi dry cake with a very small layer of solids, throwing away a perfect pre-coat layer that could have worked for a much longer cycle, not considering all the time lost in cleaning and re-pre-coating before starting the next cycle.

Autor: Alfredo Rapetta

Co-Autor: Jan Hermans

About the author

Leave a Reply