luid Mechanics
F
EGME333
December 7, 2020
 

Design project

C​ENTRIFUGAL​ P​UMP​ ​AND​ P​IPE​ S​IZING

Jacob Husband, Minjae Kim

Department of Mechanical Engineering
California State University, Fullerton
Fullerton, CA, USA

P​ROBLEM​ DESCRIPTION
A pump-piping system that can keep a filled h​ is given by
The Energy formula ​
1-million-gallon capacity water tank is wanted. There will be a
h = Δz + V 2 /2g(Σk + f * (l/d)) (2)
change in viscosity as the temperature of the groundwater that
will be filled daily is 10℃. Aquifer is 0.8 miles from the tank where Δz is the height of the tank and the rest of it is a
and 150 ft lower than the tank. Water will be used around 1.5 summation of major and minor losses. g is the gravity, f is
million gallons per day and the time to fill the tank should not the friction, l is length, and d is the diameter of the pipe.
be longer than 8 hours each day. There should be four butterfly The velocity V is given by
valves with variable openings, 10 elbows that are 90 degrees,
and galvanized iron pipe of a size should be selected in the V = 4Q/πd2 (3)
design. Pipe diameter and the impeller diameter will be where the variables are all mentioned in the Energy formula.
calculated to minimize the cost of the total price to design the The electricity cost per hour is given by
pipe.
E lectricity = 0.7457 * W * (Qmax /Q) / (0.55) (4)
PUMP-PIPE SIZING where C is the energy cost per hour, h is the head, c is the cost
The function of the pump has four variables, which include rate per kWh, μp is the pump efficiency, and μm is the motor
flow rate, head loss, diameter, and pump speed, with two efficiency.
unknowns, which are the diameter and flow rate. Flow rate is C_Q​ is given by
The Flow coefficient ​
determined by the given volume and time. Since the volume is
C Q = Q/(⍵ * D3 ) (5)
fixed, which is 1.5 million gallons, time should be modified and
tested. The flow rate of the pump must be equal to the flow rate where omega is the flow speed.
of the pipe, and the head loss of the pump must be equal to the Table 1. Intersection Point Qpipe = Qpump
sum of the head loss of the pipe and the vertical displacement
between the aquifer and the tank. Thus, the same flow rate that Case Pipe Diameter Impeller
is used for the pump will be applied to the pipe and the energy (ft) Diameter (ft)
equation. The function of the pipe also has four variables, 1 1.00 2.3
which include flow rate, head loss, diameter, and the roughness 2 1.09 2.3
of the material. The material for the pipe is given as galvanized 3 1.00 2.4
iron. The two unknowns of the pipes are also the flow rate and
the diameter. From here, trial and error is applied. For the flow
Table 2. Flow Rate and Head Loss
rate, one must change the value of time, also guess the number
for the diameter of the pipe and the impeller. Flow Rate
Case Head Loss (ft)
The Flow Rate ​Q​ is given by (ft^3/s)
1 8.03 267.98
Q = /Δt (1)
2 7.09 209.17
where ‘nu’ is the volume of the tank and Δt is a change in 3 9.53 316.34
time​.

1 Husband, Kim
 The first chart shows the Loss Coefficient and Power
Coefficient in respect to the Flow Rate Coefficient.

Figure 3. Power Coefficient, Efficiency vs Flow Rate

Coefficient

Figure 1. Loss Coefficient, Power Coefficient vs Flow Rate The fourth chart shows the Pressure head in respect to the
Coefficient Flow Rate at Peak Efficiency with omega varies, preferably
between 900 and 1800 rpm. The green curves represent the
pump loss while the red curves represent pipe loss.
The second chart shows the Loss Coefficient and
Efficiency in respect to the Flow Rate Coefficient.

Figure 4. Pressure Head vs Flow Rate when Omega varies

Figure 2. Loss Coefficient, Efficiency vs Flow Rate
Coefficient COST ANALYSIS
If the pipe and the impeller diameters are larger, then the
cost becomes larger, but capable of having quicker flow rate as
it helps decrease the time frame from 8 hours. The pump speed
varies between 900 and 1800 rpm, with a set diameter, and
chosen iterations of flow rate. Because it is 10 cents per
kilowatt-hour, the shorter time it takes, less costly it will be for
the electricity, but it requires more electricity to pump larger
flow of the water. The top of the curves for each pump diameter
is when the pump speed reaches 1800 rpm. The bottom of the
The third chart shows the Loss Coefficient and Efficiency
curves are for when the pump speed is at 900 rpm.
in respect to the Flow Rate Coefficient.

2 Husband, Kim
DESIGN DESCRIPTION
The determined pipe diameter is 1.00 feet or 12
inches, the impeller diameter is 2.3 feet or 27.6 inches, and the
flow rate is 8.03 cubic feet per second. The cost of the pump
and motor is $3,500 plus $1,500 per inch of impeller size, so
$3,500 + $1,500(27.6) = $44,900. The cost of the pipe is $1 per
inch of diameter per foot length, so $1*(12)*(4224) = $50,688.
The cost of the valves is $300 plus $200 per inch of pipe size,
so $300 + $200(12) = $2,700, but we need four valves, so the
total cost of the valves is $10,800. The cost of the elbows is $50
plus $50 per inch of pipe size, so $50 + $50(12) = $650, but we
need ten elbows, so that total cost of the elbows is $6,500.
When we plug the numbers in for the energy per hour formula,
the total electricity would cost $229.38. The total cost of the
whole project is $113,117.38. The energy cost per day of usage
is higher for the cheapest design to be created initially. After
one year, the cheapest design is pipe diameter of 1.09 feet and
impeller diameter of 2.3 feet. After five years, the cheapest
design is pipe diameter of 1.27 feet and impeller diameter of
2.4 feet. This design that is the cheapest at 5 years, will
continue to outperform the other designs because the energy
cost is the lowest.

3 Husband, Kim
 APPENDIX A

MATLAB C​ODE

close all;
clc;

% Given information
% 1-mil-gallon tank
% small model pump: D = 5.45 in, convert to ft
sample_impeller_diameter = 5.45/12;
% w = 1,760 rpm, T = 20 deg C
% efficiency peaks at 55% when
% Q = 45 gpm and ha = 23 ft
% ------------------------
% Tank begins @ 0 gal out of 1 mil
% T = 10 deg C, l = 0.8 miles,
% 0.8 miles = 4224 ft
length = 4224;
% z2 = 150 ft, z1 = 0 ft,
z2 = 150;
z1 = 0;
% -1.5 mil gal / day expended
% -62.5 thousand gal / hour expended
% -0.5 mil gal / 8 hours expended
% 8 hours to fill tank 0 gal to 1 mil
% +1.5 mil gal / 8 hours to fill
% the 1 mil gal tank @ 8 hour mark
% while it loses liquid
% 6.9625 ft^3/s is minimum Q
Q_min = 6.9625;
Q_max = Q_min * 8;
flow_iteration = (Q_max-Q_min)/100;
% 4 butterfly valves, variable openings
% 10 90 deg elbows,Kl = 0.3
loss_coefficient = 0.3;
% 4 butterfly valves
butterfly_valve = 0.4;
% galvanized iron pipe
% epsilon = 0.015 mm = 0.00042 ft
epsilon = 0.00042;
% pipe diameter ?
% impeller diameter ?
% pump speed w = 900 - 1800 rpm
% pump speed w = 94.248 - 188.495rad/s
% 1760rpm * 2pi / 60 = 184.306rad/s
omega = 184.306;
omega_max = 188.495;
omega_min = 94.248;
omega_iteration = (omega_max - omega_min)/100;
% pump scaling laws
% Qd1/Qd2 = d1^3/d2^3 - flowrate
% hd1/hd2 = d1^2/d2^2 - head
% Wd1/Wd2 = d1^5/d2^5 - power

4 Husband, Kim
% gravity
gravity = 32.2;
% density of water is 62.4 lbs/ft^3
density_water = 62.4;
% kinematic viscosity of water is 1.4081 * 10^-5 ft^2/s @ 10 C
% dynamic viscosity of water is 2.7310 * 10^-5 lbf*s/ft^2
viscosity_water = 0.00002731;
% kinematic viscosity of water is 1.0789 * 10^-5 ft^2/s @ 20 C
% dynamic viscosity of water is 2.0895 * 10^-5 lbf*s/ft^2
viscosity_sample = 0.000020895;
% Ha = delta(p)/g*dwater
[delta_P, power_in, power_sample, power_out_actual, mean_velocity,...
head_loss_actual, flowrate_sample, flowrate_coefficient, head_coefficient,...
power_coefficient] = deal(zeros(1,13));
% Reynolds number is dwater * mean velocity * diameter / viscosity

Qsample = [0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60];
head_sample = [28, 28, 29, 29, 28, 28, 27, 26, 25, 23, 21, 18, 15];
sampleeffic = [0, 13, 25, 35, 44, 48, 51, 53, 54, 55, 53, 50, 45];
efficperc = [0, 0.13, 0.25, 0.35, 0.44, 0.48, 0.51, 0.53, 0.54, 0.55, 0.53, 0.50, 0.45];

for h = 1:13
% changing units to ft^3/s
flowrate_sample(h) = Qsample(h) / (7.481 * 60);
% solving for the power output
power_sample(h) = flowrate_sample(h) * head_sample(h) * density_water;
flowrate_coefficient(h) = flowrate_sample(h)/(omega*power(sample_impeller_diameter, 3));
head_coefficient(h) = gravity*head_sample(h)/(omega*omega*sample_impeller_diameter*sample_impeller_diameter);
power_coefficient(h) = power_sample(h)/(density_water*power(omega, 3)*power(sample_impeller_diameter, 5));
delta_P(h) = head_sample(h) * gravity * density_water;
end

% let diameter of pipe = 1 ft min, 10 ft max, iterate 100 times

[mean_velocity_list, reynolds_list, f, pipe_losses, total_pipe_losses] = deal(zeros(100,100));
[pipe_d_list] = deal(zeros(1,100));
pipe_min = 1;
pipe_max = 10;
pipe_iteration = (pipe_max-pipe_min)/100;
flowrate = Q_min;
for j = 1:100
for k = 1:100
% diameter increases left to right
% flowrate increases top to bottom
% flowrate is held constant left to right
% diameter is held constant top to bottom
mean_velocity_list(j, k) = (flowrate * 4) / (3.1415 * power(pipe_min, 2));
pipe_min = pipe_min + pipe_iteration;
end
pipe_min = 1;
flowrate = flowrate + flow_iteration;
end
% with mean velocities for 100 flowrates and 100 pipe diameters
% find Reynolds values for all 10000 data points
for j = 1:100
for k = 1:100
% diameter held constant for a single column

5 Husband, Kim
 % flowrate held constant for a single row
reynolds_list(j, k) = density_water * mean_velocity_list(j, k) * pipe_min / viscosity_water;
pipe_min = pipe_min + pipe_iteration;
end
pipe_min = 1;
end
% with 10000 values of reynolds (100 diameters x 100 flowrates),
% we can solve for f using the alternative explicit expression
for j = 1:100
for k = 1:100
% diameter held constant for a single column
% flowrate held constant for a single row
v1 = epsilon/pipe_min;
v2 = v1/3.7;
v3 = v2^(111/100);
inside_log = v3 + (6.9/reynolds_list(j,k));
findinglog = (-1.8)*log10(inside_log);
f(j, k) = 1/(findinglog^2);
pipe_min = pipe_min + pipe_iteration;
end
pipe_min = 1;
end
% major and minor head losses in the pipe
for j = 1:100
for k = 1:10
% diameter held constant for a single column
% flowrate held constant for a single row
pipe_losses(j, k) = ((mean_velocity_list(j, k)^2)/(2*gravity))*(4.6 + ((f(j, k) * length)/pipe_min));
total_pipe_losses(j, k) = pipe_losses(j, k) + 150;
if j == 1
pipe_d_list(1, j) = pipe_min;
end
pipe_min = pipe_min + pipe_iteration;
end
pipe_min = 1;
end

% instantiating arrays to hold values

[omega_list, head_list,power_list, qs] = deal(zeros(100,1));
[omega_mega_list, head_mega_list, power_mega_list] = deal(zeros(100,13));
[imp_list] = deal(zeros(1,100));
[intersection_list] = deal(zeros(2,34));
col_start = 1;
% pump scaling calculations using coefficient functions for a single
% impeller diameter
begin = 1;
for j = 2:42
q = Q_min;
impD = 1 + (j-2)/10;
imp_list(j-1) = impD;
for i = 1:100
% finding omega for varying flowrate
omega_list(i) = omega_scale(flowrate_coefficient(10), q, impD);
% making sure omega > 900 or omega < 1800 rpm
if (omega_list(i) > omega_max) || (omega_list(i) < omega_min)
omega_list(i) = 0;

6 Husband, Kim
 end
% saving all omega's in array
omega_mega_list(i,j) = omega_list(i);
% finding head loss for impeller D with the omega values
head_list(i) = head_scale(head_coefficient(10),impD,omega_list(i), gravity); % saving all head loss in array
head_mega_list(i,j) = head_list(i);
power_list(i) = power_scale(power_coefficient(10), impD, omega_list(i));
power_mega_list(i,j) = power_list(i);
% saving list of flowrates
qs(i) = q;
% increasing flowrate by a small amount
q = q + flow_iteration;
end
% transposing the head loss list to fit intersection format
h1 = head_list.';
h1(h1==0) =nan;
% transposing the flowrate list to fit intersection format
q1 = qs.';
% beginning plotting for figure 5
figure(5)
% for loop t makes sure that all the intersection points are found for
% one curve of impeller diameter to all curves of pipe diameter
for t = 2:9
% this InterX function was an intersection function created online
% by NS (2020). Curve intersections (https://www.mathworks.com/matlabcentral/fileexchange/22441-curve-intersections),
MATLAB Central File Exchange. Retrieved December 8, 2020.
stinker = total_pipe_losses.';
P = InterX([q1; h1],[q1; stinker(t-1, :)]);
% saving all the intersection points in an array, flowrate in row 1
% and head loss in row 2
if P ~= 0
intersection_list(:,begin) = P(:, 1);
begin = begin + 1;
end
% columns have to be increased by 2 to hold x1 y1 x2 y2 in the
% previous 2 columns
col_start = col_start + 2;
hold on;
% plotting the intersection points with blue circles
p = plot(q1, h1, q1, total_pipe_losses(:,(j-1)),P(1,:),P(2,:),'bo');
end
% deciding flowrate1 to flowrate2 and from what headloss1 to headloss2
% to view the plot from
axis([7.5 55 150 1000]);
pump = plot(q1, h1);
pump.Color = 'g';
pipe = plot(q1, total_pipe_losses(:,(j-1)));
pipe.Color = 'r';
end
q2 = q1.';
hold off;
% creating legend, axis labels, title
title(legend, 'Legend');
h(1) = pump;
h(2) = pipe;
legend(h, {'Pump Loss', 'Pipe Loss'});

7 Husband, Kim
xlabel('Flowrate (ft^3/s)');
ylabel('Pressure Head (ft)');
title('Pressure Head vs Flowrate, Varying Omega');

[impeller_diameters, pipe_diameters] = deal(zeros(34,1));

cost = zeros(34,8);
% finding diameters via indexing
for i = 1:34
[~,flowindex] = min(abs(q1(1,:)-intersection_list(1,i)));
[~,index] = min(abs(head_mega_list(flowindex,:)-intersection_list(2,i)));
impeller_diameters(i) = imp_list(index);
[~,index] = min(abs(total_pipe_losses(flowindex, :)-intersection_list(2,i)));
pipe_diameters(i) = 1 + ((index-1)*0.09);
end
diameters = horzcat(pipe_diameters,impeller_diameters);
flowrates = intersection_list(1,:);
flowrates = flowrates.';
diameters = horzcat(diameters,flowrates);
heads = intersection_list(2,:);
heads = heads.';
biglist = horzcat(diameters, heads);

for i = 1:34
[cost(i,1), cost(i,2), cost(i,3), cost(i,4), cost(i,5), cost(i,6), cost(i,7), cost(i,8)] = cost_analysis(biglist(i,1), biglist(i,2), biglist(i,3),
biglist(i,4));
end

figure(2)
yyaxis left;
plot(flowrate_coefficient, head_coefficient);
hold on;
yyaxis right;
plot(flowrate_coefficient, power_coefficient);
title('Loss Coefficient, Power Coefficient vs Flowrate Coefficient');
xlabel('Flowrate Coefficient');
ylabel('Loss Coefficient, Power Coefficient');
hold off;
title(legend, 'Legend');
legend('Loss Coefficient', 'Power Coefficient');
figure(3)
yyaxis left;
plot(flowrate_coefficient, head_coefficient);
hold on;
yyaxis right;
plot(flowrate_coefficient, efficperc);
title('Loss Coefficient, Efficiency vs Flowrate Coefficient');
xlabel('Flowrate Coefficient');
ylabel('Loss Coefficient, Efficiency');
hold off;
title(legend, 'Legend');
legend('Loss Coefficient', 'Efficiency');
figure(4)
yyaxis left;
plot(flowrate_coefficient, power_coefficient);
hold on;
yyaxis right;

8 Husband, Kim
plot(flowrate_coefficient, efficperc);
title('Power Coefficient, Efficiency vs Flowrate Coefficient');
xlabel('Flowrate Coefficient');
ylabel('Power Coefficient, Efficiency');
hold off;
title(legend, 'Legend');
legend('Power Coefficient', 'Efficiency');

function impeller_diameter = impeller_scale(Cq, Q2, omega2)

impeller_diameter = power( Q2 / (omega2 * Cq), (1/3));
end

function omega_calc = omega_scale(Cq, Q2, D2)

omega_calc = Q2/(Cq*power(D2,3));
end

function head = head_scale(Ch, D2, omega2, g)

head = Ch*power(omega2,2)*power(D2,2)/g;
end

function powers = power_scale(Cp, D2, omega2)

powers = Cp*62.4*power(omega2, 3)*power(D2, 5);
end

function [impeller, valves, elbows, pipes, electricity, cost, cost1yr, cost5yr] = cost_analysis(d, D, q, h)
effic = 0.55;
q_max = 55.7;
impeller = 3500 + (1500*D*12);
valves = 4*(300 + (200*12*d));
elbows = 10*(50 + (50*12*d));
pipes = 1*(4224*12*d);
electricity = (448.831 * 0.7457 * q * h * 0.1 * ( q_max / q ) ) / (3960 * effic);
cost = pipes + elbows + valves + impeller + electricity;
cost1yr = cost + (electricity*365);
cost5yr = cost + (electricity*365*5);
end

9 Husband, Kim
Impeller Cost Valve Cost Elbow Cost Pipe Cost ($) Electricity ($) Total Initial Total Cost 1 Total Cost 5
($) ($) ($) Cost ($) Year ($) Year ($)

44900.00 10800.00 6500.00 50688 229.38 113117.38 196840.62 531733.58

44900.00 11664.00 7040.00 55249.92 179.04 119032.96 184381.94 445777.87

46700.00 10800.00 6500.00 50688 270.77 114958.77 213790.04 609115.12

46700.00 11664.00 7040.00 55249.92 193.90 120847.82 191622.66 474722.02

46700.00 12528.00 7580.00 59811.84 165.67 126785.51 187255.02 429133.07

46700.00 13392.00 8120.00 64373.76 151.86 132737.62 188165.23 409875.71

48500.00 10800.00 6500.00 50688 340.83 116828.83 241230.12 738835.30

48500.00 11664.00 7040.00 55249.92 214.16 122668.08 200836.08 513508.08

48500.00 12528.00 7580.00 59811.84 175.09 128594.93 192503.69 448138.70

50300.00 10800.00 6500.00 50688 482.13 118770.13 294748.55 998662.22

50300.00 11664.00 7040.00 55249.92 242.91 124496.83 213157.69 567801.16

50300.00 12528.00 7580.00 59811.84 187.15 130406.99 198717.61 471960.09

52100.00 11664.00 7040.00 55249.92 286.22 126340.14 230809.46 648686.77

52100.00 12528.00 7580.00 59811.84 202.91 132222.75 206284.03 502529.14

53900.00 11664.00 7040.00 55249.92 357.80 128211.72 258809.59 781201.05

53900.00 12528.00 7580.00 59811.84 224.08 134043.92 215831.37 542981.17

55700.00 11664.00 7040.00 55249.92 496.55 130150.47 311392.55 1036360.85

55700.00 12528.00 7580.00 59811.84 253.64 135873.48 228452.25 598767.32

57500.00 12528.00 7580.00 59811.84 297.30 137717.14 246230.16 680282.24

59300.00 12528.00 7580.00 59811.84 367.41 139587.25 273693.00 810115.97

61100.00 12528.00 7580.00 59811.84 496.90 141516.74 322886.78 1048366.93

61100.00 13392.00 8120.00 64373.76 261.63 147247.39 242742.49 624722.87

62900.00 12528.00 7580.00 59811.84 812.11 143631.95 440050.34 1625723.91

62900.00 13392.00 8120.00 64373.76 304.35 149090.11 260179.37 704536.40

10 Husband, Kim
64700.00 13392.00 8120.00 64373.76 370.86 150956.62 286318.96 827768.35

66500.00 13392.00 8120.00 64373.76 487.49 152873.25 330805.82 1042536.10

68300.00 13392.00 8120.00 64373.76 742.45 154928.21 425922.82 1509901.24

68300.00 14256.00 8660.00 68935.68 308.00 160459.68 272878.93 722555.93

70100.00 14256.00 8660.00 68935.68 369.61 162321.29 297227.39 836851.82

71900.00 14256.00 8660.00 68935.68 472.26 164223.94 336600.06 1026104.55

73700.00 14256.00 8660.00 68935.68 675.76 166227.44 412880.72 1399493.83

77300.00 15120.00 9200.00 73497.6 454.21 175571.81 341358.82 1004506.86

79100.00 15120.00 9200.00 73497.6 616.43 177534.03 402529.47 1302511.25

84500.00 15984.00 9740.00 78059.52 565.31 188848.83 395188.63 1220547.83

11 Husband, Kim