UNIDO Pump expert Harry Rosen of TAS Online talks about a recent pump assessment for a group of Abalone farmers in the Western Cape of South Africa, which involves pumping seawater at ever-changing sea levels in a suction lift application.
Click here to download and read pdf
One of the more scenic views of a pump station, the sea water intake pump station for an Abalone farm showing the suction lift conditions.
I recently held a training course on pumps with a group of Abalone farmers in Gansbaai, a drive of around 40 minutes from Hermanus in the Western Cape. Besides being one of the most beautiful parts of the country to visit, Gansbaai is world-renowned for its shark cave diving trips, which offer a chance to come up close and personal with a Great White.
This industry is going through tough times as there are no white sharks to be seen all the way down the coastline. Apparently they are the favourite day-time snack of the Orca and whole groups of these killer whales apparently swim up and down the SA coastline attacking and killing great white sharks. There are now far fewer sightings of these beasts in the waters around SA – maybe thanks to the Free Willy movie about saving an Orca whale.
Abalone ¬ known in the Western Cape as Perlemoen – is an eastern delicacy farmed commercially in SA, with most of the natural abalone having been hunted to near extinction. Abalone farming requires pumping huge amounts of seawater through hundreds of tanks filled with abalone, supplying nutrients to the small creatures as well as removing their waste. Unfortunately two characteristics of the sea make this a difficult pumping application.
The level of the sea water will generally be below the level of any pump station. This leads to a suction lift or negative suction application – excluding, of course, countries with areas of land below sea level, such as Holland, where sea water is pumped back into the sea.
The level of the sea varies greatly, both over a single day – between low and high tides – as well as over a month – during spring low tides, for example, at full and new moon when the sun and the moon align.
Suction lift applications require a level of planning and suitable engineering design to ensure that there is always a positive pressure pushing liquid into the suction of the pump. As much as I get fed up with them, pumps do not ‘suck’ in any sense of the word.
Which means that they require a minimum absolute pressure on the suction side, or they will start cavitating and end up destroying themselves.
The suction side pressure is called the Net Positive Suction Head available or NPSHA and is made up of the following three components:
1. Atmospheric pressure
- One advantage of being at sea level as there is 101 kPa or 10 m of head available to push sea water into the pump.
- Up in Johannesburg, at 1 600 m above sea level, we only have 83 kPa to help out.
2. Suction level:
- A positive help in many applications but these pumping stations are mainly located at the maximum sea level or high tide mark. This reduces the risk of flooding during high seas but it means that for the majority of time, the sea is going to be at least 2.0 to 4.0 m below the pumps. During a spring low tide, the sea level will be even lower.
- A possible solution would be vertical turbine pump, which have the pump motor out of the water, while the impeller bowl is submerged in the sea. Submersible pumps are another option, but both of these introduce further issues and complexity and were not considered.
3. Pipe Friction
- To maximise the available suction pressure we need to minimise friction losses through the suction pipe, thus each pump has its own dedicated suction pipe. For minimum friction we go for the largest diameter pipe possible with the shortest length and as few bends as possible.
- We do not install any valves on the suction pipe, even if it means leaving out the non-return or foot valve so common in suction lift applications. It is better to re-prime the pump every time it stops than to incur extra friction losses through the foot valve. Luckily, these pumps run 24 hours a day and only stop during planned maintenance or emergencies.
The vapour pressure is insignificant for water where the water temperature is less than 30 °C.
System design for abalone farming
Since suction is the dominant feature of the application, we have designed the system to optimise the NPSH available (NPSHA) at the pump’s suction. We have also installed both suction and discharge pressure gauges for each pump to show exactly what is going on with the pumps.
We selected a pump with the lowest possible NPSH requirement (NPSHR) at the duty required, which happens to be an NPSHR of 4.0 m. Including a margin of 1.0 m, we get a total positive suction head requirement of
5.0 m, or 50 kPa. So long as our NPSHA exceeds this amount, we will not expect the pumps to cavitate.
If pumping systems were this simple, though, I would probably be out of a job and have nothing interesting to write about. However things change, sea levels rise and fall, organic growth builds up inside our large diameter pipes reducing their effective diameter, kelp and seaweed lifted up through storms clog the inlets to the pumps, etc. The pumps were running fine last week but suddenly three weeks later we have problems.
Ongoing monitoring
At the time of my visit the suction pressure gauge was reading -61 kPa. This sounds serious, but what does it mean – are we pulling a vacuum, is the pump cavitating, what now?
This is where successful abalone farmers need to understand the relationship between gauge pressure and absolute pressure. NPSH is expressed in absolute pressure – which includes atmospheric pressure – whereas the suction gauge is reading gauge pressure. These are not the same thing, like comparing apples with oranges!
Pabsolute = Pgauge + Patmospheric
If we switch the terms around we get:
Pgauge = Pabsolute - Patmospheric
Our NPSH required for the pump was 5.0 m or 50 kPa absolute pressure. If we subtract the 101 kPa of atmospheric pressure, this gives us gives us -51 kPa gauge pressure. Now we have something useful to work with – if the gauge pressure on the suction side of the pump is less than -51 kPa, then our pump is in trouble.
But our suction gauge is reading
-61 kPa, so our pump must be cavitating: Right? Wrong! The suction gauge was installed 2.0 m above the level of the suction pipe or pump centreline.
This gauge height correction must be added back to the suction pressure reading as the liquid loses pressure as it rises in the tube connected to the gauge. The gauge pressure at the pump suction is therefore -41 kPa, and even though it is way less than zero – or what many people would call pulling a vacuum – our pump is not cavitating.
This is the power of a compound pressure gauge on the suction side that can read negative pressure. If we used a standard discharge gauge on the suction it would probably read zero all the time, not telling us anything useful.
If the gauge pressure dropped below
-50 kPa we know our pump is close to cavitation and we need to stop the pump, fix the suction conditions and only then restart the pump. A pump cavitating under these conditions could destroy the impeller in a matter of weeks.
What were the recommendations?
- Work out the NPSHR for the pump – from the original pump curve and data sheet. Add 1.0 m for a margin, and convert from absolute pressure to gauge pressure.
- Move the suction gauge in line with the pump. If this is a problem with respect to access and being able to view the gauge easily, reduce the NPSHR value by the gauge height – in this case this would reduce NPSHR to -71 kPa.
- Make sure this value is clearly shown in the pump station, by marking up the pressure gauges with red and green zones as shown in the Figure 1. This clearly shows when the pump begins to experience suction problems.
- When the gauge reads less than the minimum or inches towards the red zone, then something must be done. This could be happening for a number of reasons:
- Over time, friction increases through the suction pipe due to organic growth. The solution is to clean out the pipe, using either a high pressure wash or manually.
- Spring low tides cause unusually low suction levels, causing the suction pressure to drop into the red. Stop the pump at these times. It only happens twice per month for a couple of hours. Build additional storage into the system to allow for this and other times like load shedding when you cannot run your pumps
- Regularly clean out the strainers and cages around the pump inlets. Once again, the suction needle inching towards the red zone will tell you when this needs to be done.
As with all difficult pumping applications, the starting point is designing the system and selecting the pump that best suits the site-specific conditions. In this case the negative suction head and low NPSHA was key. But a well-designed system is not enough: conditions vary over time so it is essential to continuously monitor pump operation. This will prevent the pumps from destroying themselves, even under the harshest unforeseen circumstances.
And finally, a negative pressure on the suction side is not necessarily a problem. This happens in many applications with a positive suction head that we are not even aware of. It is only when this value falls below the absolute pressure required by the pump, the NPSHR, that we have a real problem.
