If an oily water mix is placed into a tank then speration of the two parts will begin with the lighter element rising to the top. The rate the seperation occurs is governed by several factors including the difference in specific gravities and the force of gravity acting upon it.
For mixes placed into a settling tank there is little that can be done about the gravity but the difference in the specific gravities can be increased by heating. This because water density changes at a much reduced rate when compared to oil. The limiting factor to this is that the water cannot be heated above 100'C for obvious reasons.
A wide shallow tank will increase the rate of clarification over a tall thin tank
When a volume of light oil is placed into a tank contain a weir and a quantity of water the fluids will tend to arrange themselves as shown above. The height of the water in the weir rises to a point governed by the volume ( and thereby relative height) and specific gravity of the light oil.
Knowing this it is possible to form a rudimentary purification system.
As a oil/water mix is fed into the tank seperation begins with heavy particulates falling to the base of the tank along with water which joins the other water excess overflowing the heavy phase weir. Hopefully clear oil passes over the light phase weir. The problem arises that to ensure their is suffcient time to allow for full(seperation of the oily mix the flow would have to be very small relative to the size of the tank.
Fluid moving between two plates has a velocity greatest at mid point and minimum approaching the plates.
a particle entering into the plates will tend to be pushed upwards by the fluid flow. All the time centrifugal foce tends to retard the horizontal component of the movement causing the partical to approach the underside of the top disc. As it approaches the fluid flow velocity reduces. The centrifugal force eventually overcomes the force acting on the partical due to fluid movement and th epartical starts to move towards the oute rim. The centrifugal force acting on a particle is proportional to its mass therefore a small particle will tend to move further under the influence of fluid flow. Indeed a particle small enough will be carried through the plates and out with the discharge. In this way it can be seen that reducing the flow rate to a purifier will tend to increase the quality of the output.
The basic centrifuge differs than that described above most obviously by sitting on its side. In reality it takes the form of a round bowl a cross section of which will show something like that seen above. Gravity is replaced by centripetal force as the bowl is spun at high revolutions thereby creating very high g-forces.
A disc stack is incorporated to encourage a laminar flow increaseing improving the seperation effect. Dirty oil is introduced via a centreline oil feed dip tube. The oil is led to distribution holes which are refected in the disc stack but not the dam
The graph shown above is one typical of one found in a purifer instruction book for selecting appropriate gravity disc size. Shown on the diagram is an example of an oil of sg 0.93 at 0'C. The sg at 15'C for use with this graph is found by projecting along a horizontal line to 15'C. This step would be omitted if the sg at 15'C was already known. A line is then drawn parallel to the pre-drawn sloping lines. Where the drawn sloping line cuts the appropriate oil supply temperature isothermal then This becomes the selection point for the disc. This is found simply by ascertaining which size band the point lies in.
The majority of purifers found on board are of the self cleaning type intat they are able to open the bowl to discharge any accummulated sludge. Apart from the sliding bowl the main difference is the centripetal pump over the simple design. In this a fixed centrigual style impeller is mounted in the light phase outlet drawing the oil and discharging it at pressure sufficient to deliver it to the receiving tank. A discharge valve is fitted which is adjusted to give a constant back pressure in the bowl. The adjustment of this back pressure tends to move the position of the interface but more importantly increases the oil in the light phase delivery chamber increaseing the immersion depth of the lip of the pump. This reduces possiblility of air being entrained and removes foaming.. In the event of bowl failure back pressure will fall, this may be detected by a pressure switch initiaing a shut down
Considerable torque would be required to direct drive the bowl upto speed using an appropriately sized electric motor. In addition very high loading would occur on the gear train, to prevent scuffing due to oil film breakdown would require large mating areas therby large gear trains which would again increase the starting load.
A centrifugal clutch arrangement is fitted which has between 2 and 6 ferrodo lined brake pads. These are designed to slip during the start up period and also to a much lesser extent during the speed up period after de-sludge. Purifer manufacturers will usually quote a maximum and MINIMUM start up time. As the pads wear it may be necessary to remove and restore the mating surface to keep the start up time correct. As a last measure the number of pads should be altered
The electric motor may be of special design allowing for a long period of slight overload during the start up period.
The gear train is generally a single stage worm and wheel arrangement with the wheel being made of a softer material. Lubrication is normally splash only, the viscosity of the oil is essential to prevent wear as the form of lubrication is mainly boundary therefore the wear is governed by the viscosity and additives contained within the oil.
When wear occurs it will be scuffing and relative movement between the mating faces polishes out any pitting. As wear worsens galling occurs destroying the running surface. This damage is reflected in both elements therefore both should be changed.
As well as overload other causes of premature failure are poor design ( step forward westfalia), poor material choice, poor lube oil choice, too long a de-sludge period relative to supplied oil quality, out of balance bowl, failing bearing set in particular the vertical shaft upper resilient bearing arrangement
The use of planned maintenance is essential particularly with respect to bearing changes. It is strongly recommended to
monitor condition using vibration analysis
Shown is a typical circuit for a lube oil system although it can equally be applied to a fuel system. Control is achieved by the three way valve which eitherdiverst oil to recirculate or sends it to the purifer.
Oil flow rate is controlled by the oil control valve situated before the positive displacement delivery pump which is driven off the purifier horizontal shaft via a weak link arrangement
Back pressure from the purifer is controlled at oultet via the back pressure control valve
Shown is typical damage to the sealing face of a sliding bowl. This has been caused by either poor assembly or by hard material being trapped aft the bowl closes. Unfortuately it is more likely to be the former.
Failure is detected by loss of sealing water as seen down the sludge shute. In addition there will be carryover to the heavy phase and loss of discharge pressure as the seal is exhausted.