9 replies to this topic

[kkala]

Mind goes directly to cavitation; there will be stronger evidence for it, if pump flow is verified to fall considerably from the specified value of 1500 m3/h (also resulting in discharge pressure reduction). Screw pump cavitation seems more "progressive" compared to centrifugal pumps (see http://www.springer...2mt38646581631/), yet symptoms are similar.
-Assuming HFO sg = 1.0, NPSHr=5.4 m CL in pump data sheet; suction pressure = 1.41 bara during test (14.4 m CL). What was operating temperature of (non heated) fuel oil during test? Vapor pressure of "dry" fuel oil is usually quite low (to ensure flash point>60 oC), but vapor pressure of any water content is additive to it, since the two liquids are not miscible. Moreover HFO temperature should be at least 90 oC to suspect cavitation from water vapor (vap press = 0.7 Bara) under test conditions, and this 90 oC is probably higher than operating temperature. Anyway check both water and HFO vapor pressures at actual operating temperature.
-Can you feed the screw pump with another liquid (water at ambient temperature?) to see whether it runs smoothly? You can exclude pump mechanical problem in this way.
-At the end I would repeat the test using fuel oil with more instruments (at any case additional suction manometer for verification of 0.4 Barg) in the presence of a representative of pump supplier.
-HFO heating to a higher temperature (i.e. reducing suction frictional ΔP) is not certain to solve the problem, according to my understanding.

Edited by kkala, 01 September 2011 - 01:59 AM.

[JMW]

So without heating, what temperature is the fuel?
380cst is the viscosity at 50C.
At 20C: 4,650cst (approx, based on CIMAC data for a 390cst fuel)
At 30C: 1800cst
At 40C: 790cst....
Heating is important.
Especially for the intended 680cst....
AT 20C: 10,200cst
At 30C: 3,600cst
at 40C: 1500cst
It isn't just cavitation due to the suction problems, but if you do get the discharge line primed the headloss may cause the pressure relief valve to trip so you will be recirculating fuel outlet to inlet.

[kkala]

-Low fluid vapor pressure (even if oil contains water at 35 oC) and suction pressure of 0.4 Barg makes the possibility of cavitation remote, especially if flow remains at 1500 m3/h during test. During a new test run, try to check suction pressure (with additional manometer) and to measure flow rate, directly (e.g. by a meter) or indirectly (e.g. by measuring level in destination tank, etc). Actual flow rate is of interest, to be compared to 1500 m3/h (specified in the attached hydraulic calculations for linesizing).
-It is noted that heavy fuel oils stored here (s.g. =0.95-0.98) would require continuous heating to (say) 55 oC; handling temperature of 35 oC (lower in the night) seems too low. But suction pressure of 0.4 Barg (as measured) is in excess of what you need. So measurement of suction and discharge pressure during test run has also to be continuous and reliable.
-Mentioned cleaning and priming of suction line are judged to be on the right direction.
-I think fuel oil gums (rather from polymerization / oxidation) are formed in long time; in your case fuel oil can be locally viscous (or even agglomerated?) due to moderate cold, and this could remain during day; but a screw pump is generally tolerant to it; and suction line is not plugged, since it has enough suction pressure at rated flow. So ambient 15-20 oC at night may not have a significant result, on the condition that suction pressure remains at 0.4 barg during test run.
-Probably it is good to call an engineer of the pump supplier to solve the problem, if it continues. Will the supplier be present during the pump commissioning? This is normal practice here, especially for pumps of such size.

[kkala]

Eager to know the result of this case. Test run can give a guidance for solution, watch pressures continuously (and flow rates, if possible).
Concerning event of PSV chattering, I suppose you could detect whether noise comes from pump or PSV, if you approach the pump discharge closely.
I remember plant operators able to detect cavitation from the characteristic noise of the specific pump, in the common case when cavitation produces noise. So probably an operator could diagnose PSV chattering from the noise.
Water mixed with fuel oil is not expected to cause cavitation at 35 oC operating pressure.

Edited by kkala, 04 September 2011 - 05:10 AM.

[JMW]

Heavy fuel oils are today more usually blends of high viscosity residuum and low viscosity distillates rather than straight run distillation products.
While cavitation may not be a problem, flashing off of volatiles may be.

Since the client didn't include the heaters, did they also neglect insulation? What ever heat the fuel has ought to be preserved as best as possible though, to be fair, insulation is often more about protecting operators from hot pipes than it is to conserving heat or energy.
I know of cases with bitumen installations where local cols spots caused complete blockages. In instances where there was not proper protection the pump simply kept running and ballooned the flow meter till the bolts sheared.

If this isn't continuous flow, do they flush the lines with diesel or reverse flow the pump to empty the pipes back into storage?
Starting cold with a flooded pipe can be an issue both for suction and discharge. I have come across some installations where they just couldn't get any reasonable flow from a cold start (and in one installation they used flexible couplings which simply collapsed).

Now, if the discharge is partially blocked, the pressure relief bypass will cut in. But if the suction is starved? Kkala?

PS.You quoted the density as 930kg/m3 at 50C which is 950.4kg/m3 at 15C which fits with expectations and the range of values Kkala mentioned.

[kkala]

One can hardly find straight run atmospheric residues today; cracked fuel oils are much worse in plugging (cannot be used in reciprocating internal combustion engines here). But any blockage at suction line would lower its suction pressure, then flowrate and discharge pressure. Same in case of starvation of supply in suction line. Suction pressure (and also flow rate, discharge pressure), should be closely and continuously monitored in next test run, to diagnose whether blockage / starvation occurs, even instantaneously, or not. Suction pressure of 0.4 barg does not indicate cavitation, but this should be checked.
Volatiles could cause even cavitation by increasing fuel oil vapor pressure, but normally these are controlled indirectly to keep fuel oil flash point below 66 oC.

Edited by kkala, 05 September 2011 - 04:04 PM.