Experiment No 06

To determine the operating characteristics of two-way centrifugal pump in parallel combination

6.1 Introduction
6.1.1 Parallel combination of pumps:
Pumps running in parallel will increase flow (Q), but not head (H). Pumps are operated in parallel
when two or more pumps are connected to a common discharge line and share the same suction
conditions. Water pressure boosting applications use, for example, pumps from the CR range to
ensure steady flow through a common discharge line. When using the Grundfos sizing tool for the
product in question, you can size identical pumps either in parallel or as single pumps in series.
If pumps are running in series, this will result in increased head (H), but not flow (Q). [1]

6.1.2 Two pumps in parallel combination

Parallel pumps are useful for systems with considerable discharge variations and with no appreciable
head change. In parallel, each pump has the same head. However, each pump contributes to the
discharge so that the total discharge is equal to the sum of the contributions of each pump. Thus, for
pumps:
Q = Q1 + Q2 + Q3 ... + Qn
The composite head characteristic curve is obtained by summing up the discharge of all pumps for
the same values of head. A typical pipe system curve and performance point of the pumps are shown
in Figure. [2]
6.1.3 Working of two pumps in parallel combination
In fluid systems, connecting pumps in parallel offers a significant advantage over a single pump
setup. This configuration increases the overall flow rate by essentially creating separate channels for
the fluid each pump contributes its individual flow rate, resulting in a combined flow that is the sum
of the two. It's important to note that the pressure (or head) in the system remains relatively constant
compared to a single pump. This is because both pumps discharge into a common area, and the
pressure equalizes throughout. Parallel pumps are ideal for situations demanding a higher flow rate
than a single pump can provide, such as chilled water systems in buildings, process plants with
fluctuating flow requirements, and municipal water supply applications. However, for optimal
performance, it's recommended to use identical pumps in parallel. Dissimilar pumps can lead to
uneven flow distribution and inefficiencies. Additionally, check valves at each pump inlet are often
used to prevent backflow when a pump is turned off.

6.1.4 Advantages and Disadvantages

6.1.4.1 Advantages:
 Increased Flow Rate: The primary benefit of connecting centrifugal pumps in parallel is
the significant rise in flow rate. Each pump acts as an independent channel for the fluid, and
their combined flow rate becomes the sum of the individual capacities. This is crucial for
applications requiring higher flow than a single pump can handle.
 System Redundancy: With two pumps operating, if one pump fails, the other can maintain
some level of flow, allowing for continued operation or a controlled shutdown. This
redundancy enhances system reliability, particularly in critical applications.
6.1.4.2 Disadvantages:
 Constant Pressure: While flow rate increases, the pressure (head) in a parallel configuration
remains practically the same as a single pump. This is because both pumps discharge into a
common header, and the pressure equalizes throughout the system. If higher pressure is
required, series configuration might be a better option.
 Complexity and Cost: Installing and maintaining two pumps requires more space, piping,
and control systems compared to a single pump. This translates to higher initial and ongoing
costs.
 Uneven Flow Distribution: For optimal performance, the pumps should be identical.
Significant differences in size or performance can lead to uneven flow distribution. One
pump might end up doing most of the work, reducing efficiency and potentially damaging the
underutilized pump. Check valves at each pump inlet can help mitigate this to some extent.
[3]

6.1.5 Applications
Centrifugal pumps are a popular choice for parallel configurations due to their inherent ability to
handle increased flow rates without a significant pressure penalty. Here are some applications where
two centrifugal pumps in parallel prove beneficial:
 High Flow Rate Systems: In applications requiring a substantial volume of fluid movement,
such as municipal water supply, parallel pumps become crucial. A single pump might not be able
 to meet the peak demand of a city, whereas parallel pumps allow the system to deliver the
necessary flow while maintaining pressure for residential and commercial use.
 Chilled Water Systems in Buildings: Large buildings often utilize central air conditioning
systems with chilled water loops. Paralleling centrifugal pumps ensure adequate cool water
circulation throughout the building, especially during peak cooling periods. This maintains
consistent comfort levels for occupants.
 Process Plants with Fluctuating Flow Demands: Many industrial processes involve variations
in flow rate requirements. Chemical plants, for instance, might experience periods of high flow
during specific stages of production. Having parallel pumps allows the system to adjust to these
fluctuations by turning pumps on or off as needed. This flexibility ensures efficient operation and
avoids unnecessary energy consumption by a single oversized pump.
 Redundancy and System Backup: Paralleling pumps introduce redundancy into a system. If
one pump malfunctions, the remaining pump can keep the system operational at a reduced
capacity, providing time for repairs or maintenance without complete system shutdown. This is
crucial in applications like fire sprinkler systems where maintaining water pressure is critical
during emergencies.

6.2 Literature Review

6.2.1 Analytical Review
The maximum operating efficiency of a centrifugal pump is different when it is parallel operated
with different pumps. To study the reduction in the maximum operating efficiency of pumps in
parallel operation compared with their rated efficiency, 28 pairs of pumps with the same rated head
and different rated flows were tested orthogonally in parallel operation. A determination of the
pumps operating flow and head was done to calculate the maximum operating efficiency in different
combinations. The results showed that the maximum operating efficiency of a pump in parallel
operation was higher than 85% of rated efficiency when the flow ratio of two pumps was within 1.6.
When the flow ratio was increased to 2.0, the maximum operating efficiency of the pump will
sharply decline to 70% of rated efficiency or less. Therefore, the operating efficiency a pump can
keep is 70% or more under the condition that the flow ratio between the pumps is less than 2. [5]
P¿ =VI
Po =ρg H T Q

H s =H 3−H 1 + H d
 H T =H 2−H 1

6.2.2 Modeling and simulation

This article focuses on developing simulation models for large-scale irrigation systems that utilize
multiple centrifugal pumps operating in parallel, all powered by a single photovoltaic (PV) generator.
It highlights the importance of such models in optimizing water flow and energy usage within these
systems.
The key takeaway is that simulating parallel pump configurations offers advantages over simply
multiplying the performance of a single pump by the number of pumps (N). This article explores
these advantages and presents models that account for factors like:
 Individual pump characteristics
 Available PV power
 System dynamics
By incorporating these aspects, the models can predict the optimal combination of pumps and their
operating frequencies to maximize water flow based on the available power from the PV generator.

To depict the pump power characteristics of radial centrifugal pumps, a physical model was
developed. The model relies on established empirical equations. To parameterize the model for
specific pumps, physically interpretable tuning factors were integrated. The tuning factors are
identified by using the Levenberg–Marquardt method, which was applied to the characteristic curve
at a constant speed. A cross-validation of the physical model highlighted the advantage of
representing the set of performance curves with less deviation compared to approximation functions.
Calculating the entire set of performance curves requires only one pump characteristic curve at a
constant speed. Therefore, only a single measurement is necessary. Furthermore, the physical model
can be used to calculate the changes in the set of performance curves due to pre-whirl. This increases
the accuracy of flow rate calculations when pre-whirl occurs. [6]
In the present study, we intend to investigate the effect of the number of blades on pump efficiency.
Additionally, we want to choose the most efficient turbulence model for the problem. Considering
the various turbulence models, three known turbulence models of standard k–ε, RNG k–ε and RSM
are utilized at a constant blade number of 6 for the pump. In order to show the results in a more
practical order non-dimensional parameters defined in reference are used and for the sake of
generality the data are reported in non-dimensional form.
Fig shows the head coefficient vs. flow coefficient for the three turbulence models compared with
available experimental data. The curves show that with increasing the flow coefficient, the head
coefficient is decreased. Comparing various turbulence models data with experimental data, it has
been concluded that each of these turbulence models provide acceptable results, but two models
RNG k–ε and RSM show better agreement than the standard k–ε model.

Fig. Head coefficient vs. flow coefficient with three different turbulence models and one available
experimental data.
The power coefficient and efficiency curves vs. flow coefficient are plotted in figures respectively.
As shown in fig, in low and high mass flow rates different turbulence models approach each other.
In fig, in addition to low and high mass flow rates, in mid mass flow rates, difference between
various turbulence models are low because efficiency considers the effect of both head and power.

Fig. Power coefficient vs. flow coefficient with three different turbulence models.
 Fig. Efficiency vs. flow coefficient with three different turbulence models.

6.2.3 Experimentation
Large amounts of energy is being used in a wide range of applications to transport liquid. This paper
proposes a generic solution for minimizing power consumption of a generic pumping station
equipped with identical variable speed pumps. The proposed solution consists of two sequential
steps; firstly, the energy consumption for a pre-selected set of active pumps are formulated as a
convex optimization problem and that can be solved analytically. Secondly, the problem of choosing
the number of active pumps are formulated and solved using a convex solver subject to constraints.
The proposed solution is compared with a conventional affinity-law based solution. The
experimental analysis showed that the proposed solution and method lead to more precise pump
modeling and simplicity in solving this type of pump optimization problem.
An experimental setup has been used for the experiments and validating the results in this paper. The
Piping and Instrumentation Diagram (P&ID) of the pumping station is shown in Fig. This testing
facility consists of three identical variable speed pumps, each with differential pressure and flow rate
measurements. To emulate different hydrodynamic resistances downstream a control valve is
deployed. The flow meters are magnetic flow meters from Emerson, and the pumps are Grundfos
CRE 5–5, all equipped with VFDs. The nominal design flow rate for all the pumps are 6.9 m3/hr at a
head of 32m. For power measurements, a Zimmer power analyzer LMG670 is used.

This study investigated optimal scheduling of a group of identical variable speed pumps in terms of
minimizing energy consumption subject to a satisfactory pump performance. The proposed model
structure was compared to affinity-based models and the proposed model provided a significant
improvement in prediction accuracy, especially at lower speeds.
The proposed solution consists of two sequential parts; firstly, the optimization problem with respect
to pump speed, given a set of operating pumps and expected operation point in head and flow rate,
was formulated. It was concluded that under these conditions, the optimal solution was running the
pumps at identical speed. Secondly, the optimal number of active pumps was formulated as a convex
problem with constraints and solved using a standard solver.
The proposed solution presumes the performance curves remain constant over time; the scheduling
can therefore be solved offline. However, in practice is not the case. The model could be extended to
cover non-identical pumps in combination with performance map adaptation for taken changing
behavior into considerations, such as cavitation and wear. If performance map adaptation is
deployed, the optimization must be done online.
The proposed scheduling method was tested and compared to an alternative method. The resulting
energy reduction from the proposed method is as much as 30%. [7,8]

6.3 Methodology
6.3.1 Experimental Setup
 Hydraulic Bench
 Two centrifugal pumps in parallel combination
 Stopwatch
 Input and output pressure head gauges
 System head gauge
 voltmeter

Ammeter

Output head
RPM meter

Input head

System Head
Flow control
valve

6.3.2 Procedure
 Set up closed-loop test bench with two centrifugal pumps, piping, gauges, flow meter, valves,
and data acquisition system (optional).
 Connect pumps in parallel, with both discharges feeding a common header.
 Install pressure gauges at the inlet of each pump and one at the common discharge header.
 Place a flow meter at the common discharge header to measure total flow rate.
 Connect the system to a reservoir and adjust valves to control flow.
 Close all control valves, fill system with working fluid, bleed air, ensure electrical connections
and safety.
 (Optional) Run individual pump baseline test collecting head-flow data for each pump.
 Start both pumps in parallel, adjust control valve for different flow rates.
 Record discharge pressure from the common header gauge and total flow rate for each setting.
  Repeat for various flow rates including near zero and maximum.
 Analyze data: plot head-flow curves for each pump individually (if baseline data exists).
 Analyze how combined flow rate changes with head compared to single pump.
 Gradually close control valve, stop pumps, drain working fluid.

6.4 Observations and Calculations

Motor Datu
Volum Tim Input Outpu Manifol Voltag Curren Total Efficienc Power Power
Rpm Flowrate m
e e Head t Head d Head e t Head y Output Input
Speed Head

% Watt Watt
rpm m3 s m3/s m m m V A m m
s s
0.527
3000 0.005 4.22 1.18x10-3 -0.075 0.52 0.248 210 1.5 0.8 27% 84 315
5
0.546
3000 0.005 6 0.83x10-3 -0.068 0.54 0.320 210 1.5 0.8 20% 61 315
8
0.582
3000 0.005 6.46 0.77x10-3 -0.025 0.58 0.413 210 1.5 0.8 19% 60 315
5

6.5 Result and Discussion

Efficiency vs flow rate

30%
25%
20%
efficiency

15%
10%
5%
0%
0.7 0.8 0.9 1 1.1 1.2 1.3
flow rate x10^-3

Output Power vs Flow rate

100
80
Output Power

60
40
20
0
0.7 0.8 0.9 1 1.1 1.2 1.3
Flow rate x10^-3
 Flow rate vs Total Head
0.6
0.58

Total Head
0.56
0.54
0.52
0.5
0.48
0.7 0.8 0.9 1 1.1 1.2 1.3
Flow rate x10^-3

6.6 Conclusion
6.7 References
1. Pumps in Parallel https://www.engineeringtoolbox.com/pumps-parallel-serial-
d_636.html#:~:text=When%20two%20or%20more%20pumps,one%20pump%20can
%20handle%20alone.
2. HOW TO OPERATE CENTRIFUGAL PUMPS IN SERIES OR PARALLEL Jorie Ballun,
January 22, 2015 https://blog.craneengineering.net/operating-centrifugal-pumps-in-series-or-
parallel
3. Pumps in parallel https://www.grundfos.com/solutions/learn/research-and-insights/pumps-in-
parallel
4. Parallel Pumps: Reliability & Efficiency Issues
https://www.waterworld.com/home/article/16189751/parallel-pumps-reliability-efficiency-
issues
5. Yi Wen, Xiaochu Zhang, Pu Wang, Volume 132, Issue 3, June 2010, The Relationship
Between the Maximum Efficiency and the Flow of Centrifugal Pumps in Parallel Operation
6. Pawel Olszewski, Volume 178, 15 September 2016, Pages 527-539, Genetic optimization
and experimental verification of complex parallel pumping station with centrifugal pumps
7. Roberto Alves de Almeida & Augusto Nelson Carvalho Viana, 26 April 2008 Volume 1,
pages 167–173, Optimization of parallel variable-speed-driven centrifugal pumps operation
8. Jamal Arafeh and Pawel Olszewski, Volume 231, 1 December 2018, Pages 1146-1158,
Parametric analysis of pumping station with parallel-configured centrifugal pumps towards
self-learning applications