The pump unit operation in HYSYS has an NPSH page on the rating tab. So, you might think that a quick glance will give you valuable information with little or no effort. You'd better think twice! Let's start with the things you would probably think of yourself, but it might be useful to get a reminder.
In most cases your pump will draw its feed from a tank or separator. so things to consider are:
1 - Pressure in the tank
2 - Elevation difference between the bottom of the tank and the pump
3 - Level of liquid in the tank
4 - Pressure drop in the piping from the tank to the pump
If you only model the pump and not the upstream equipment, it is up to your manual calculations to make sure you account for all the above to get the pump feed pressure correct. If you have modelled the upstream equipment, you'll have to pay attention to a few things to get this right:
1 - In steady state, HYSYS does NOT account for static head, except for pressures inside the pipe segment. So you need to use something to account for the liquid level in the tank. This could be a valve with a negative pressure drop that is calculated based on the liquid level or it could be a vertical pipe with a length that matches the liquid level.
2 - Use a pipe segment to represent the piping from the tank to the pump. You can also account for fittings here. It is easy to add an extra bit of pipe here to account for the liquid level.
Your model would look like the picture below.
So, having paid attention to all the details, it seems like the NPSH for the pump is high enough. But if you know that in the pictured example the liquid level is about 6 m above the inlet nozzle of the pump and the fluid is n-Butane close to the bubble point, an available NPSH of 11 m seems suspiciously high. So what is wrong?
HYSYS considers the pressure of the feed stream of the pump to be the pressure of the flowing liquid at the pump inlet. In Flarenet terms this is the static pressure or also the pressure you would measure with a pressure sensor that has its measuring surface parallel to the direction of flow. So, the NPSH calculation adds the velocity head to that pressure. To calculate this velocity head it uses the pump inlet nozzle size. The default nozzle size which is defined on the Rating tab, Nozzles page is 50 mm.
Even a small change in diameter will have a really big impact on the results. In my example, changing the nozzle size from 50 mm to 60 mm makes the NPSH drop from 10.98 m to 5.13 m. Changing the diameter to 100 mm results in an NPSH of only 0.39 m.
Also note that changing the nozzle diameter in steady state will not trigger a solve and hence the NPSH result will only be updated after something else forces HYSYS to resolve.
In most cases your pump will draw its feed from a tank or separator. so things to consider are:
1 - Pressure in the tank
2 - Elevation difference between the bottom of the tank and the pump
3 - Level of liquid in the tank
4 - Pressure drop in the piping from the tank to the pump
If you only model the pump and not the upstream equipment, it is up to your manual calculations to make sure you account for all the above to get the pump feed pressure correct. If you have modelled the upstream equipment, you'll have to pay attention to a few things to get this right:
1 - In steady state, HYSYS does NOT account for static head, except for pressures inside the pipe segment. So you need to use something to account for the liquid level in the tank. This could be a valve with a negative pressure drop that is calculated based on the liquid level or it could be a vertical pipe with a length that matches the liquid level.
2 - Use a pipe segment to represent the piping from the tank to the pump. You can also account for fittings here. It is easy to add an extra bit of pipe here to account for the liquid level.
Your model would look like the picture below.
In dynamics, HYSYS can account for static head, but remember this is an option that is turned off by default and you'll have to set the elevation of ALL pieces of equipment correct.
So, having paid attention to all the details, it seems like the NPSH for the pump is high enough. But if you know that in the pictured example the liquid level is about 6 m above the inlet nozzle of the pump and the fluid is n-Butane close to the bubble point, an available NPSH of 11 m seems suspiciously high. So what is wrong?
HYSYS considers the pressure of the feed stream of the pump to be the pressure of the flowing liquid at the pump inlet. In Flarenet terms this is the static pressure or also the pressure you would measure with a pressure sensor that has its measuring surface parallel to the direction of flow. So, the NPSH calculation adds the velocity head to that pressure. To calculate this velocity head it uses the pump inlet nozzle size. The default nozzle size which is defined on the Rating tab, Nozzles page is 50 mm.
Also note that changing the nozzle diameter in steady state will not trigger a solve and hence the NPSH result will only be updated after something else forces HYSYS to resolve.
1 comment:
very useful tips
Post a Comment