11 replies to this topic

[Edavan]

Can anyone share their advice..?

Vertical pumps (1st pump) which takes liquid from sump and discharges to the suction of the second pump (horizontal).

Both the pumps have same capacity but with different discharge head.

Please advice:

1) Any specific thing we need to consider for pump in series operation;

2) If the first pump discharge head is on the higher side; whether the first pump will feed the enough capacity to the second pump.

Any suggestions on considering the capacity and discharge head of the first pump?

[ankur2061]

Vinprab,

The head for pumps in series is the sum of individual heads with the capacity remaining the same. Have a look at this link:
http://www.tpub.com/.../h1012v3_76.htm

Hope this helps.

Regards,
Ankur.

[kkala]

Not having experience on the topic, I can only tell an opinion to start discussion. I assume there is a good reason for two pumps in series, but please explain why you cannot place only one vertical (not submerged?) pump. I assume you design a new system.

1. Pumps (including vertical pumps) can practically supply a definite minimum head (say 25m Liq, depending on manufacturer). We find difficulty in finding pumps of lower head, so you can specify the vertical pump based on rather low head, but well covering NPSHa of the horizontal pump at any possible case.
2. Vertical pump should be simple, in the usual operating range and if possible of good quality. Remember that it is not so easy to carry this pump to the maintenance shop.
2. Rest head required can be taken by the horizontal pump, design flow rate of both pumps should be same.
3. It is noted that shutoff pressure in the discharge of horizontal pump is the sum of the shutoff pressures of the two pumps.
4. Min return flow from the discharge of horizontal pump is needed. There must be a valve at the suction of horizontal pump (i.e. discharge of vertical), so min flow return from the discharge of vertical pump is also needed. Probably this additional return line of vertical pump may not be installed if pumps start up only locally (judged to depend also on liquid handled, e.g. water, naphtha, etc).
5. Check valve at horizontal pump discharge is usually needed.

Edited by Art Montemayor, 29 December 2009 - 09:07 PM.

[akslzf]

Pull out the specification sheets of both pumps & check the following:

1) Normal & minimum flow of first pump shall be at least 10% above the minimum continuous stable flow (MCSF)of the 2nd pump.
2) When pumps run in series, their heads are to be added. Therefore, max allowed working pressure (MAWP) of the 2nd pump must be at least 5% above shut off pressure of 2nd pump
3) Whether the mech seals of the 2nd pump (and the first) are adequately designed to handle the pressure
4) Does the first pump (Booster pump) have adequate NPSHa? Else, if that starves, the 2nd pump will also cavitate.
5) Trip logic to be provided to trip 2nd pump, in case booster pump trips, for the same reason mentioned in 4 above
6) Pipeline & fittings' rating shall be adequate enough to handle the 2nd pump's discharge pressure

BTW also consider shear properties of fluid & its temperature. Above all, draw a head-flow curve for 2-pumps-in-series operation, duly considering the system resistance to see if there's really an advantage in running them in series!!

Edited by Art Montemayor, 29 December 2009 - 09:08 PM.

[kkala]

"Pull out the specification sheets of both pumps & check the following
1) Normal & minimum flow of first pump shall be at least 10% above the minimum continuous stable flow (MCSF)of the 2nd pump.
2) When pumps run in series, their heads are to be added. Therefore, max allowed working pressure (MAWP) of the 2nd pump must be at least 5% above shut off pressure of 2nd pump
3) Whether the mech seals of the 2nd pump (and the first) are adequately designed to handle the pressure
4) Does the first pump (Booster pump) have adequate NPSHa? Else, if that starves, the 2nd pump will also cavitate.
5) Trip logic to be provided to trip 2nd pump, in case booster pump trips, for the same reason mentioned in 4 above
6) Pipeline & fittings' rating shall be adequate enough to handle the 2nd pump's discharge pressure
BTW also consider shear properties of fluid & its temperature
Above all, draw a head-flow curve for 2-pumps-in-series operation, duly considering the system resistance to see if there's really an advantage in running them in series!!"

Indeed useful guidelines, in case two existing pumps are to be connected in series. A few clarifying questions in reference to the above.
2) Max shutoff pressure (Pmax) of 2nd (horizontal) pump will be: 1st pump max shutoff pressure + 2nd pump shutoff differential pressure (also considering static levels). Assuming no surge, Pmax can be considered as MAWP / design pressure of 2nd pump casing (and discharge piping), why the 5% margin over Pmax? Is it applicable by some practice or code?
3) Shall design pressure (DP) of mechanical seal be same as DP of pump casing? Once we observed in ordered pumps that mechanical seal DP was lower than max shutoff pressure = casing DP. The manufacturer supported that mechanical seal DP can be about half the casing DP, due to pump rotation (centrifugal force?). Clarification on this point of DP by members would be useful.
- fluid shear properties: that is viscosity for Newtonean fluids?

[akslzf]

1) Shutoff pressure of 2nd will NOT be: booster pump shutoff pressure + 2nd pump shutoff pressure, for the simple reason that, if booster pump is shutoff, there's no flow to 2nd pump. It is therefore the sum of rated discharge head of the 2nd pump in SO condition + rated discharge head of first pump. (convert head to pressure). The 5% margin is just a good practice, keeping in mind, changes in SG (mainly for hydrocarbons) that could affect the 2nd pump's discharge pressure

2) Seal chamber pressure (depending on centrifugal pump volute/diffuser design & impeller design) can be approximated to be suction pressure + 10-50% of head developed by pump. Therefore, it is going to be highest during SO condition. Opening pressure of seals, as far as I've seen, is usually in the range of 20-25 bar, whereas, pump casing MAWP is much higher, to take care of discharge pressure conditions.

[fallah]

Shutoff pressure of 2nd will NOT be: booster pump shutoff pressure + 2nd pump shutoff pressure, for the simple reason that, if booster pump is shutoff, there's no flow to 2nd pump.

I think when 2nd pump would be in SO condition dictates this condition (no flow) to 1st pump due to series operation's nature.
Thus,Shutoff pressure of 2nd will be: booster pump shutoff pressure + 2nd pump shutoff pressure

Edited by fallah, 30 December 2009 - 02:17 AM.

[narendrasony]

I think when 2nd pump would be in SO condition dictates this condition (no flow) to 1st pump due to series operation's nature.
Thus,Shutoff pressure of 2nd will be: booster pump shutoff pressure + 2nd pump shutoff pressure

First (booster) pump will be running on minimum flow in that case. So, I think Shutoff pressre for 2nd pump should be : Booster pump discharege head at minimum flow condition + 2nd pump shutoff pressure.

[kkala]

I think when 2nd pump would be in SO condition dictates this condition (no flow) to 1st pump due to series operation's nature.
Thus,Shutoff pressure of 2nd will be: booster pump shutoff pressure + 2nd pump shutoff pressure

First (booster) pump will be running on minimum flow in that case. So, I think Shutoff pressure for 2nd pump should be : Booster pump discharege head at minimum flow condition + 2nd pump shutoff pressure.

1. Our intent is to specify design pressure (or MAWP) for pump casing and downstream piping (same for suction piping according to several practices).
2. If a valve at 2nd pump discharge is closed, minimum flow is expected to operate. But this is a control action not recognized as "reliable enough" (like e.g. a PRV) by several practices.
API complies with this concept (specific reference cannot be given due to my vacations from work). E.g. we would not consider pump minimum flow for sizing a relief valve at the discharge piping, blocked outlet case.
3. I have some doubt on the above only for the case of mechanical recirculation valves (type yarway or equivalent), but at any case we do not consider minimum flow at work (general practice) for the specification of design pressure. This is conservative, yet difference between operating pressure at min flow and shutoff pressure is not great for the plurality of pumps (indicated by their curves).

[Edavan]

Thanks for the valid inputs. Please also advice the following

1) Rated/normal capacity should be same for both (1st and 2nd) pumps (or)First pump rated/normal capacity should be slighly higher than the second pumps

2) Suction pressure of second pumps is considered as 1.0 bar(g), whether this is adequate for series operation.

[S.AHMAD]

Thanks for the valid inputs. Please also advice the following

1) Rated/normal capacity should be same for both (1st and 2nd) pumps (or)First pump rated/normal capacity should be slighly higher than the second pumps

2) Suction pressure of second pumps is considered as 1.0 bar(g), whether this is adequate for series operation.

Hi,
1). Rated capacity of both pump should be the same and no need to over design the 1st pump. Of course nothing wrong to over design but not necessary in this case.

2). Whether 1 bar is sufficient or not depends on the NPSH required by the second pump. So you need to check the NPSH required and decide what is the discharge pressure of the 1st pump shall be. However, take into consideration worst case scenario whereby suction pressure of the first pump is at lowest possible pressure i.e lower liquid level of source (storage?) and possible increase in flow rate in the future (higher demand than anticipated).

Regards

[kkala]

"Whether 1 bar is sufficient or not depends on the NPSH required by the second pump. So you need to check the NPSH required and decide what is the discharge pressure of the 1st pump shall be. However, take into consideration worst case scenario whereby suction pressure of the first pump is at lowest possible pressure i.e lower liquid level of source (storage?) and possible increase in flow rate in the future (higher demand than anticipated)"

One addition to the valid remark by S.Ahmad: We have found difficulty in finding (refinery) pumps of head lower than (say) 25 m Liq in the market, so it may be preferable to specify a differential pressure of (say) 2.5 bar for the vertical pump and the balance of ΔP for the subsequent horizontal pump (see post Dec 23rd).