Water supply system steep vs flat curve 3

[desann]

Hi. this is a follow up on my previous post. I think it would be better to make a new thread because there is a clear, specific question now.

My project is about supplying water to our fogging system which is basically another pump and also end user flow.

The requirements from the device's manufacturer are 12 m3/hour at 3-4 bar. However real flow at which is the system operating is 4 or 8 m3/hour, depending on whether single string is on or both. Please note, that the flow is always restricted to 4 or 8 m3/hr by the system.

I would like to use 2 pumps in series of which the second pump is supposed to be Ebara either Matrix or 3M. First pump will be submerged in the water tank, supplying Ebara which is supposed to act as a pressure booster. The supply line will be regulated by VFD and pressure control loop. There will be a pressure tank and high flow filter unit in the system.

Please find below our system curve along with the pump characteristics. The dotted lines, barely visible are standalone pumps, the bold lines are pumps in series and system curves.
I created system curves for 3, 3,5 and 4 bar that is a range required for the end user. Also I created the characteristics for 10 and 20% speed reduction.

I can see that the Matrix pump has a much steeper line than 3M. By looking more closely I would say by going for the "steep" pump it will need more precise speed tuning but I can get the output want.

Also important to mention,the steep one is significantly cheaper.

I would be very interested in getting a more detailed view what are the real advantages and disadvantages of both solutions and which one fits our system better.
Due to the lack of practical experiences I cannot predict that, so I would like to ask you for advice. Is it all about the VFD setting and fine tuning in my scenario or do I miss something?







Thank you.

 

[goutam_freelance]

he requirements from the device's manufacturer are 12 m3/hour at 3-4 bar

But according to diagram above the flow must be twice the displacement pump flow, so 16 m3/h (or 8 m3/h if one pump is operating).

The critical aspect here is the maintenance of NPSHA of DP, so the fogging system supplier is suggesting an overdesign of the secondary pump. This is reasonable as the fogging system will run only occasionally.

If you have an additional pump for the double flow condition, the additional pump must have a stabilized flow before switching on the 2nd DP, otherwise there may be momentary cavitation in the DP suction side.

Additionally, I did not see any requirement for constant pressure at DP suction(1-2 bar is mentioned). I suggest discussing with the fogging system supplier whether this can be slightly more, in the range of 3-5 bar, such that the requirement of VFD can be obviated.

 

[bimr]

Your pressure is relatively high. Consider adding some type of pressure relief so that you don't over-pressurize the piping.

 

[desann]

But according to diagram above the flow must be twice the displacement pump flow, so 16 m3/h (or 8 m3/h if one pump is operating).

The critical aspect here is the maintenance of NPSHA of DP, so the fogging system supplier is suggesting an overdesign of the secondary pump. This is reasonable as the fogging system will run only occasionally.

If you have an additional pump for the double flow condition, the additional pump must have a stabilized flow before switching on the 2nd DP, otherwise there may be momentary cavitation in the DP suction side.

Additionally, I did not see any requirement for constant pressure at DP suction(1-2 bar is mentioned). I suggest discussing with the fogging system supplier whether this can be slightly more, in the range of 3-5 bar, such that the requirement of VFD can be obviated.

That is the thing, there are two information

• fog system supplier requires 1,5*rated flow = 12 m3/hr or 6 m3/hr at 3-4 bars
• PD pump manufacturer 2*rated flow = 16 or 8 m3/hr at 1-2 bar

I asked the supplier why this mismatch and he told me that during the designing the system they hired professional pump engineer which gives them those figures. Also they pointed out that our fog system is operating only on 90-100 bar but the pump is rated 120 bar. But I don't know, if that information is valid, or just an assumption. It's because with decreased pressure at the output of fogging system pump the flow also decreases. I don't see any point why the equation of 2*flow is not supposed to be vaid in any scenario.

Regading the secondary pump and swiching of 2nd DP, the ramp up of PD takes some time, and my idea is creating an analogue signal between VFD1 and VFD2, if the VFD 1 reaches specific frequency, VFD 2 will go in.

 

[LittleInch]

The requirement for 2 x PD pump capacity is just ensuring there is always sufficient flow to avoid any PD pump pulsation or velocity head issues on the inlet side. It's a bit lazy, but then as they are not supplying the feed pump they really don't care and know this will be more than enough capacity.

In general centrifugal pumps can be thought of as near constant pressure units (within 20%) over a wide flow range.

Your bigger pumps are a bit odd in not doing this and have a steep curve.

 

[Snickster]

I like your selection of two pumps to give the total flow required. It provides at least half of flow if one pump should go down, operates at good point on the curve and allows for the maximum flow required by the fog skid supplier.

So with the Ebara Matrix pumps it looks like you would need two size 5-4/0.9 in parallel for a head of 37 m at 4 m3/hr and 30 m at 6 m3/hr. This is coupled with a Leo 150A with 14 m at 4 m3/hr, 12 m at 8 m3/hr, and 7 m at 12 m3/hr. This will give a total of:

51 m at 4 m3hr, 49 m at 8 m3/hr and 37 m at 12 m3/hr which is within the range required by the fog skid supplier.

I agree with goutam in that the booster pumps should be activated and system pressurized before PD pumps are turned on, and also I believe that the PD pumps should be shut down before the booster pumps are shut down to prevent starving the PD pumps. An interlock with time delay is required so that PD pumps start up and shut down after a time delay after the booster pumps start up/shut down.

I don't see why you need VFD drives for the booster pumps. The flow is being controlled by the fog skid PD pumps to an exact value. This causes the two pumps in parallel and the sump pump to operate on the curve corresponding to the flow of the PD pumps, at a set head as indicated above. Why would you need further control of pressure or flow with VFD?

 

[desann]

What about if I control the feed system only with the pressure loop?

The PD pumps have ramp-up time, I can let supply pumps to ramp faster and start on slightly higher pressure set point. Also I suppose that the pressure tank will help to maintain enough water when the fog system switches on.

Fog system is operated by climate computer by various external conditions, such as solar radiation, relative humidty, temperature etc. When the condition is met, the computer sends a digital signal to on/off the fogging system. Since there is no pre-run, if I create for example 5s delay to start the feed pumps first, I will loose 50% of the fog pulse if fog pulse is supposed to be 10s.

Why VFD control

1. Soft starts/stops don't forget, fog system operates in pulses, and it's like 10s-60s fogging, 20s-120s waiting time
2. Without VFD I must link PD1 with Ebara1 and PD2 with Ebara 2, since fogging is oprating in unpredictable cycles, this will further increase number of starts of both Ebara pumps. In case of vfd control, I can just set p set point for Ebara1 and Ebara 2 will be controlled by the realized frequency of Ebara 1. Let's say If ebara1 is above 45Hz, the Ebara2 goes in, if it drops below 20 Hz, Ebara 2 goes out. (this is just a illustration idea, the correct figures will be determined by observation) This way I ensure that wheter fog1 or fog2 is running there is just Ebara1 running, and Ebara2 kicks in only if both fog pumps are operating simultaneously.

Your bigger pumps are a bit odd in not doing this and have a steep curve.

They are multistage centrifugal pumps, with 5 impellers.

 

[goutam_freelance]

hey are multistage centrifugal pumps, with 5 impellers.

It is odd to have 5 stages for around 40 m head pumps.

 

[ortanjames]

thank you for sharing this information

 

[Snickster]

I don't know enough or have enough information on the system to advise on detailed controls. I would discuss with supplier of the fogging skid as to what type of controls are normally provided on force feed pump water supply system. A guideline I always use in design is to keep things simple as possible. It reduces cost, operational and maintenance issues, and system complexity..