You are on page 1of 16

Engineering Applications of Computational Fluid

Mechanics

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tcfm20

Performance assessment of lift-based turbine for


small-scale power generation in water pipelines
using OpenFOAM

Ghada Diab, Mohamed Elhakeem & Ahmed M.A. Sattar

To cite this article: Ghada Diab, Mohamed Elhakeem & Ahmed M.A. Sattar (2022) Performance
assessment of lift-based turbine for small-scale power generation in water pipelines using
OpenFOAM, Engineering Applications of Computational Fluid Mechanics, 16:1, 536-550, DOI:
10.1080/19942060.2021.2019129

To link to this article: https://doi.org/10.1080/19942060.2021.2019129

© 2022 The Author(s). Published by Informa


UK Limited, trading as Taylor & Francis
Group

Published online: 07 Feb 2022.

Submit your article to this journal

Article views: 1168

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=tcfm20
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS
2022, VOL. 16, NO. 1, 536–550
https://doi.org/10.1080/19942060.2021.2019129

Performance assessment of lift-based turbine for small-scale power generation in


water pipelines using OpenFOAM
Ghada Diaba , Mohamed Elhakeemb,c and Ahmed M.A. Sattar d

a Civil Department, German University in Cairo, Cairo, Egypt; b Civil Engineering Department, Abu Dhabi University, Abu Dhabi, UAE;
c Department of Civil Engineering, College of Engineering, University of Tennessee, Knoxville, TN, USA; d Department of Irrigation and
Hydraulics, Faculty of Engineering, Cairo University, Giza, Egypt

ABSTRACT ARTICLE HISTORY


In-pipe water turbines have begun to gain interest for harvesting power on a small scale from Received 6 October 2021
pipe networks. However, few studies have addressed the feasibility of installing spherical lift-based Accepted 11 December 2021
helical-bladed turbines in a water supply network. Points such as the pressure drop and generated KEYWORDS
power remain unexplored. In this study, a three-dimensional numerical model, based on OpenFOAM, Gorlov; 3D numerical model;
is used to investigate the performance of the spherical lift-based helical-bladed in-pipe water tur- in-pipe turbine; helicity;
bine. The study aims to evaluate how the geometric properties of this turbine affect its performance pitch angle; turbine power;
in terms of power and efficiency, and the hydraulics of pipe flow in terms of pressure drop. The study OpenFOAM; lift-based;
considers a turbine of diameter 600 mm, with the following geometric properties: number of helical vertical axis; hydropower
blades 3, 4 and 5; blade chord length 10%, 15% and 20% of turbine diameter D; blade helicity 0°,
60° and 120°; and pitch angle −6°, −3°, 0° and 3°. These parameters are analyzed at tip speed ratios
(TSRs) of 2, 3 and 4. The results show that the five-blade turbine yields a power of 1300 W, while
the three-blade turbine yields only 870 W, at an optimum TSR of 3. A change in the chord length of
50% (from 0.10D to 0.15D) increases the turbine power by 88.4% and efficiency by 40% for the same
TSR. A further change to 0.20D gives no significant improvement in efficiency or power output. An
increase in the helical angle from 0° to 120° results in a 22.8% reduction in turbine power. The turbine
achieves the maximum power output of 1350 W at zero pitch angle, with a corresponding efficiency
of 27%. The maximum head loss observed is 1.6 m, which represents 2.7% of the total head in the
pipe. Solidity has a more pronounced effect on head losses than helicity and pitch angle.

Introduction applications include the installation on domestic pressur-


Hydropower is the most abundant, cheap and predictable ized pipelines; e.g. at the locations of pressure relief valves
source of renewable energy. Therefore, it has registered in water mains, to power water metering and water mon-
a steady and fast growth driven by the demand for itoring/control stations that are isolated from the elec-
clean, reliable, steady and affordable power. Turbines trical grid. These in-pipe turbines have diameters in the
are used to extract the potential energy of water flow range of 82–250 mm, flow velocity of 1.5–3 m/s, number
and convert it into electrical energy. While various types of blades from six to 20 and power output of 88–480 W.
of water turbines have been around for hundreds of Pipe turbines are classified according to their axis of
years, more recently, interest has grown over in-pipe rotation, into vertical and horizontal axis turbines. Verti-
turbines for harnessing clean renewable energy from cal axis turbines have the advantages of operating inde-
excess head in domestic water pipelines. Power output pendent of incident flow direction and achieving higher
ranges from 0.1 kW to 100 kW can be achieved for one efficiency at lower speed. Based on the driving force, ver-
turbine, although more power can be produced using tical axis turbines are either drag or lift based. A few in-
multiple installations. Their applications mainly fall pipe turbines of drag type have been introduced and stud-
between two types of pipelines system, namely, gravity- ied (e.g. Chen et al., 2013; Hasanzadeh et al., 2021; Jiyun
fed and pressurized pipelines. They can be installed et al., 2018; Ma et al., 2018; Samora et al., 2016). How-
in gravity-fed pipelines supplying water from elevated ever, these studies showed that the power output (and
tanks or mountains; in such cases, the free potential hence power coefficient) of drag pipe turbines is relatively
energy is converted into electrical energy. However, other low, with significant corresponding pressure losses. For

CONTACT Ahmed A. Sattar ahmoudy77@yahoo.com


This article has been republished with minor changes. These changes do not impact the academic content of the article.
© 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 537

instance, a pressure head loss of 5 m was reported by Ma parameters on turbine performance (torque, power and
et al. (2018) with a 250 mm turbine, yielding only 480 W, efficiency) for a practical pipe diameter (600 mm) under
while for a smaller diameter of 100 mm, Hasanzadeh confined flow in pipes; (2) evaluating how altering the
et al. (2021) optimized a drag-based turbine to produce operating conditions [tip speed ratio (TSR) and veloc-
200 W for the same pressure head loss of 5 m. Drag tur- ity] and geometric parameters affects the pressure drop
bines generally require more blades to build sufficient inside the pipeline under steady rotation and stall con-
flow resistance, and this has negative consequences on ditions; and (3) determining the impact of the turbine
performance and results in more flow area blockage. On shaft and hubs on the overall turbine performance, and
the other hand, lift-based turbines, developed in 1926 and the extent of the wake generated by the turbine. Rele-
known as the Darrieus wind turbine, have lower obstruc- vant variables include geometric properties such as the
tion to flow than drag-based turbines, offering generally shaft, hubs, number of blades, chord length, solidity,
better starting capabilities and a higher power coefficient. helicity and pitch angle. While most of the literature
In 2011, a commercial lift-based in-pipe turbine on pipe turbines analyzed small turbines with diame-
(Schlabach et al., 2011) was introduced to the market ter ≤ 250 mm, this study investigates a larger diameter
for generation of power from gravity-fed large pipelines of 600 mm, which represents typical main pipelines with
with diameters of 600 to 1500 mm, with power of 14 flow velocity greater than 2.0 m/s.
and 100 kW, respectively. Although a decade has passed A three-dimensional (3D) numerical model, based on
since the development of the first commercial lift-based OpenFOAM, was utilized to investigate the performance
in-pipe turbine, very few studies exist in the literature of one of the available in-pipe water turbines, namely
that evaluated the performance of this type of turbine the lift-based helical-bladed turbine. Various geometric
(e.g. Tahadjodi Langroudi et al., 2020; Purdue ECT Team, properties and operational parameters of the helical-
2016; Yang et al., 2019). Tahadjodi Langroudi et al. bladed turbine are explained first, followed by a brief
(2020) presented theoretical and numerical analyses for description of the numerical model used in the analysis
the effect of multiple design parameters on the power of the turbine, and the built mesh and boundary con-
coefficient, pressure coefficient and cavitation probabil- ditions. The model was validated against experimental
ity as performance metrics. Yang et al. (2019) investi- measurements for similar turbines available in the litera-
gated the start-up performance and pressure head loss ture. Once the model had been validated, it was used to
of lift- and drag-based turbines inside a pipe of 200 mm examine the effect of altering the turbine geometric prop-
via numerical and experimental analyses, and concluded erties and operational parameters on its performance in
that the pressure head loss is less than 3% of the total terms of turbine power, torque, efficiency and head losses.
head. The Purdue ECT Team (2016) studied the rela-
tionship between power output, diameter and head loss
in lift-based pipe turbines. Several studies have studied Methodology
lift-based turbines; however, little is known about how
Vertical axis lift-based pipe turbine
spherical helical lift-based turbines fit inside a pipe, in
terms of quantifying the head loss caused by these tur- Evolving from the Darrieus wind turbine, the Gorlov
bines and disturbance to the flow, which is crucial from helical turbine (Gorlov, 1995) was developed by chang-
a hydraulics aspect. The power output is also an impor- ing the blades from straight to helical in an attempt to
tant aspect, to make sure that the turbines satisfy the main decrease the encountered pulsatory torque issues and
purpose of installation in off-grid locations in the water increase turbine efficiency. Further, the Gorlov helical
supply network. The aforementioned studies layout the wind turbine was utilized in pipelines by changing its
foundation for a better understanding of the feasibility of shape from cylindrical to spherical (Schlabach et al.,
spherical lift-based in-pipe turbine utilization; however, 2011).
some turbine performance metrics remain unexplored, Figure 1 shows the Gorlov pipe turbine, which has a
such as the extent of the turbine wake effect on the veloc- spherical shape to fit in a pipe. The turbine shaft or axis
ity profile in the pipeline; the interaction between turbine of rotation is always perpendicular to the flow, regard-
parts, such as the hubs and shaft, with the flow of water; less of its direction; thus, it will always rotate in the same
and the pressure head loss and power associated with direction, with an angular velocity ω. The turbine blades
larger pipe diameters, and 600 mm in particular. helically sweep the sphere’s outer imaginary surface area
Therefore, this paper aims to add to the existing liter- and are connected to the bottom and top hubs. All blades
ature on lift-based in-pipe turbines, by (1) assessing the have the same cross-section, with the shape of a hydro-
performance of a vertical-axis water pipe turbine to pro- foil; usually, a NACA 0020 hydrofoil is used for better
vide insight on the influence of various geometric design hydrodynamic characteristics (Talukdar et al., 2018). The
538 G. DIAB ET AL.

Figure 1. (a) Gorlov pipe turbine components; (b) drag and lift forces on a single-blade hydrofoil.

hydrofoil chord length, c, is about 15% of the turbine swept area, and can be expressed as:
diameter, at zero blade curvature to pitch angle ratio, β.
Nc
The Gorlov turbine concept is based on the movement σ = (2)
of the hydrofoils in order to change the apparent direction πD
of the flow or Azimuth angle θ, relative to the hydrofoil and the TSR is defined as follows:
axis, and thus change the (apparent) “angle of attack” of
ωD
the hydrofoil, α. This generates lift and drag forces on TSR = (3)
each of the turbine hydrofoils. With lift force prevailing, 2U
the resultant force on each hydrofoil propels the turbine where N = number of rotor blades, c = hydrofoil chord
in the clockwise direction, causing a torque from which length, U = approaching flow velocity, ω = angular
the energy is generated. In the spherical turbine, blades velocity and D = pipe diameter. The recommended
have a helical angle, , which has an important effect value of σ for vertical axis wind turbines is in the range
by distributing evenly the hydrofoil sections through- of 0.2–0.6. For low TSR values, the turbine will tend to
out turbine rotation cycle. Thus, there will be always a slow or stall, while for high TSR values, the turbine will
hydrofoil section at every possible angle of attack. In this rotate too quickly to extract power and will be stressed
way, the sum of the lift and drag forces on each blade and may structurally fail. Thus, the optimal TSR value
does not change abruptly with rotation angle, generating corresponds to the highest power output from the tur-
a smoother torque curve, so there is much less vibration bine, and it depends on the turbine rotor and blades and
and noise than in the Darrieus design. It also minimizes is specific for each turbine.
stress in the pipe section where the turbine is installed
and facilitates self-starting of the turbine. The Gorlov
turbine can achieve efficiency of about 25–30%, where Model description
efficiency is defined as: Computational fluid dynamics (CFD) models have
shown to be effective tools not only for the preliminary
T.ω design, geometric optimization and performance assess-
η= (1)
ρgQ H ment applications of water turbines (Jiyun et al., 2018),
but also for applications such as liquid sloshing in reser-
where T = torque, ρ = water density, g = acceleration voir tanks (Ghalandari et al., 2019) and the computation
due to gravity, Q = volumetric flow rate and H = head of drag and lift coefficients (Salih et al., 2019). In this
losses across the turbine. study, a 3D numerical model, based on OpenFOAM, has
Two parameters play an important role in the perfor- been utilized to evaluate how the geometric properties of
mance of the considered pipe turbine; namely, the tip a lift-based helical-bladed turbine affect its performance.
speed ratio (TSR) and rotor solidity (σ ). The solidity is OpenFOAM (2019) is open-source software that
defined as the ratio between the total blade area to the solves the 3D unsteady Reynolds-averaged Navier–Stokes
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 539

solved in 3D. Transient simulation with adjustable time


steps, with an average of 2−5 s per time step, was imple-
mented using a sliding mesh technique. Initial runs with
approximately 1.5, 2.1, 4, 5 and 5.7 million grids were car-
ried out to study mesh dependence and optimize its size
while probing the turbine’s output power, efficiency and
head loss (Figure 3a). The results showed that the out-
put turbine power started to stabilize with a maximum
difference of 1% starting from the 4 million-grid mesh.
Thus, the 4 million grid is utilized in this study, as it
offers a good balance between computational time and
Figure 2. (a) Domain components of the test case; (b) cross- numerical accuracy.
section of the computational grid. A detailed sensitivity analysis was performed for time
step and convergence criteria (Rezaeiha et al., 2018).
Analysis showed that an azimuthal increment angle of
(URANS) equations for incompressible flow. The shear 0.0295° per time step would be sufficient for proper
stress transport (SST) k-ω turbulence model is chosen for solution convergence, with a mean Courant number
numerical system closure, since it is capable of predict- of 0.0025. The instantaneous torque output versus the
ing flow separation and adverse pressure gradients and azimuthal position during a complete turbine revolution
has been classified as one of the most reliable turbulence for 15 revolutions of five turbine cases with various rota-
models for turbine numerical simulation, having good tion speeds (from 134 to 268 rpm) was monitored. It
agreement with experimental data (Alemi et al., 2015; was found that a minimum of three turbine revolutions
Gourdain, 2015; Roy & Saha, 2013). was needed to reach a statistically steady-state solution.
To alleviate the effects of boundary conditions on Figure 3(b) shows a sample of the torque produced from
turbine simulation, inlet and outlet pipe lengths are two consecutive turbine rotations; beyond the third rev-
set 12 and 10 times the pipe diameter, respectively. olution, the torque curves overlap. The computational
The pipe diameter is considered to be 600 mm, with time required for one the cases was about 12 h to reach the
5 mm clearance between each turbine hub and pipe; steady-state solution, which was used as the initial solu-
thus, the turbine diameter is 590 mm. Turbine model tion for the transient case. The transient case solution for
geometry was first generated using SALOME (2019) three revolutions of the turbine took an average of 648 h
3D computer-aided design software. Afterwards, the (27 days) using an Intel Xeon Quad CPU E3-1271v3 @
model is imported to OpenFOAM and split into mul- 3.60 GHz. A total of 28 runs was simulated using two
tiple components, as shown in Figure 2(a) (inlet pipe, workstations for almost a year. Table 1 summarizes the
turbine and outlet pipe) containing patches for phys- turbine parameters used in different simulations of vari-
ical boundary conditions. The model mesh has a sta- ous runs. The examined parameters are the flow velocity,
tionary domain (pipe) and a rotating domain (tur- TSR, number of blades, chord length, helicity and pitch
bine) (Figure 2b). The turbine components, i.e. blades, angle.
hubs and shaft, are assigned different patches. They are
included in the spherical rotating volume, with a free
stream spherical interface boundary patch. A constant Model validation
volumetric flow rate is assigned at the pipe inlet and The model was validated using the experimental data of
a pressure head at the pipe outlet downstream of the Samora et al. (2016), examining a horizontal-axis in-pipe
interface. drag turbine, and Yang et al. examining a vertical-axis
in-pipe lift turbine. To compare the power and head pre-
dictions of the model against the experimental data, the
Results
geometry and flow conditions of the denoted experi-
A basic structured background 3D hex mesh was gen- ments were simulated using OpenFOAM. Figure 4 shows
erated using blockMesh to represent the boundaries of the predicted values of power and head from the model
the computational domain. Then, the snappyHexMesh simulation compared to the measured values from the
utility was used to generate the final hexahedra and split- experiments under various volumetric flow rates. Aver-
hexahedra mesh from the triangulated surface geome- age errors between the measured and predicted values of
tries. Because of the asymmetric nature of the turbine, turbine power and head drop were 4% and 8%. Tahad-
the mesh was generated over the whole domain and jodi Langroudi et al. (2020) reported a numerical model
540 G. DIAB ET AL.

Figure 3. (a) Error percentage versus mesh size (6 million mesh is the reference); (b) torque variation for the second and third turbine
revolutions.
Table 1. Parameters of various simulated turbine cases.
Flow velocity Chord length Helicity Pitch angle
Run serial U (m/s) TSR No. of blades N c (mm)  (°) (β)
1–3 2 2, 3 and 4 3 87 60 0
4–6 2 3, 4 and 5 4 87 60 0
7–9 2 2, 3 and 4 5 87 60 0
10–12 2 2, 3 and 4 5 58 60 0
13–15 2 2, 3 and 4 5 87 0 0
16–18 2 2, 3 and 4 5 87 120 0
19–21 2 2, 3 and 4 5 87 60 3
22–24 2 2, 3 and 4 5 87 60 −3
25 2 3 5 87 60 −6
26–28 1.5, 2.5 and 3.5 4, 2.4 and 1.7 5 87 60 0
Note: TSR = tip speed ratio.

Figure 4. OpenFOAM model results compared to experimentally measured turbine power and head loss versus volumetric flow rate.
[Experimental results from Samora et al. (2016) and Yang et al. (2019).]

error of head drop from 6% to 8%. It is observed that the these errors. However, the model was able to capture the
model tends to underestimate the head drop for high flow trend in power and head drop. Thus, OpenFOAM and
rates by up to 32% and overestimate the power for low the proposed SST k-ω turbulence model were deemed
flow rates by up to 15.7%. The approximation made in the appropriate to describe and solve the turbine numerical
geometry of the turbine prior to modeling contributed to scheme.
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 541

Figure 5. Turbine driving forces and output torque during a complete turbine revolution: (a) lift and drag forces on a single blade; (b)
torque of each blade and overall turbine torque; (c) torque and angle of attack for a single blade; (d) torque fluctuations for various flow
velocities. TSR = tip speed ratio.

Turbine torque separation starts to occur around the blade (Figure 6),
leading to a deterioration in its hydrodynamic perfor-
Lift and drag forces were calculated on a single blade
mance, by reducing the lift force approaching a stall
in a turbine (N = 5, c = 58 mm,  = 60°, β = 0° and
condition. At mid-turbine revolution, θ = 180°, lift force
TSR = 2, 4) through a complete turbine revolution, as
becomes zero and starts to change sign and thus direc-
shown in Figure 5(a). Using the OpenFOAM results, the
tion, while drag force decreases to a minimum value, then
pressure forces on each blade in the x- and y- directions
maintains the same direction and sign. Thus, for this tur-
were used to calculate both tangential, FT , and normal,
bine, 22° is the critical angle of attack of the blade, where
FN , forces on each blade, as follows:
the highest turbine torque is typically achieved during a
FT = −Fx cos(θ) − Fy sin(θ) (4) single revolution (Figure 5b). Changing the TSR from 2 to
4 causes the stall condition to occur earlier in the turbine
FN = Fx sin(θ) − Fy cos(θ) (5) revolution at a lower angle of attack, and the average blade
Following the calculation of the tangential and normal torque output increases by 150% (Figure 5c). The impact
forces on the blades, the lift and drag forces can be of changing the TSR from 2 to 4 on lift and drag forces is
calculated using the following equations: more pronounced during the first half of turbine rotation
(0° ≤ θ ≤ 180°) since it doubles both lift and drag forces;
FD = FN sin(α) − FT cos(α) (6) however, it diminishes during the second half of turbine
rotation (180° ≤ θ ≤ 360°).
FL = FT sin(α) + FN cos(α) (7)
It is expected that higher flow velocities would
The above-mentioned forces were calculated for each enhance turbine performance significantly but also cause
blade at multiple sections along the blades’ helical route. higher head losses. An increase in the flow velocity from
The performance of the turbine blades is similar to that 1.5 m/s to 3.5 m/s yielded a torque 4.5 times higher than
of a typical hydrofoil blade; the lift force on the blade is the initial value (Figure 5d). In addition, the turbine
typically much higher than the drag force during a tur- power reached up to 3600 W from 400 W (nine times
bine revolution (Figure 5a). Lift and drag forces increase higher). However, higher flow velocity induced more
with the turbine revolution (increase in azimuth angle, head losses at the turbine, which were increased from
θ), until a maximum value is reached at θ = 72°, corre- 0.6 m (at U = 1.5 m/s) to 1.6 m (at U = 3.5 m/s), an
sponding to α = 22°. At the azimuth angle of 72°, flow increase of 167%. A pressure drop of 1.6 m represents
542 G. DIAB ET AL.

Figure 6. Pressure (Pa) and velocity (m/s) distribution around a single blade (at mid-height) at different azimuth positions; tip speed
ratio (TSR) = 2.
2.7% of the total pressure head in the simulated case; the life and performance of the pipe and turbine. It was
this observation agrees with the conclusion by Yang et al. found that a flow velocity of 2 m/s would yield the highest
(2019) for a similar turbine with a smaller diameter of efficiency for the turbine cases studied in this paper with
200 mm, which induced head losses of less than 3% for relatively low torque fluctuations. Thus, in the following
the flow rates included in their study. analysis, the flow velocity was set to 2 m/s and the TSR
It is to be noted that an advantage of the turbine is that was varied by changing the turbine speed (rotations per
it does not obstruct flow in the pipeline in stall condition, minute, rpm). Preliminary runs were made for the tur-
nor does it induce high pressure losses. Figure 7 shows the bine with TSR ranging from 0.2 to 6 to find the optimal
velocity distribution along the center section of the pipe TSR value, which corresponds to the maximum output
for a stall turbine with high velocity U = 3.5 m/s. Despite power and efficiency. The simulations showed that the
the high flow velocity, the pipe velocity profile started to optimal TSR value was about 3; therefore, a range of TSR
get rid of the blade wake effects and became fully devel- from 2 to 4 was deemed appropriate to carry out the next
oped at a distance equal to around four times the pipe runs to identify the optimal number of blades and chord
diameter, with a head loss of 0.67 m. length.
Simulations showed that high flow velocity increases
the turbine efficiency from 18% (at U = 1.5 m/s) to a
maximum value of 28% (at U = 2.5 m/s). Not only does Blade number
the efficiency not improve with the increase in flow veloc-
ity beyond 2.5 m/s, but also it decreases, owing to the The effect of the number of turbine blades, N, on the tur-
increase in head losses at the turbine. As discussed, the bine performance was also investigated. Turbines with
increase in flow velocity through the turbine would yield three, four and five blades were examined. As demon-
a corresponding increase in torque and thus turbine strated by Lucid Energy Technologies (Schlabach et al.,
power. However, this increase in power has to be consid- 2011), turbine blades are fixed from the top and bot-
ered with an associated increase in head losses through tom to hubs, which cannot accommodate more than five
the turbine, and thus a decrease in the turbine efficiency; helical blades from a design viewpoint; this geometric
and a corresponding increase in the turbine torque constraint makes it more feasible to control turbine solid-
fluctuations (Figure 5d), which would structurally affect ity via chord length adjustment. Thus, the maximum
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 543

Figure 7. Velocity distribution (m/s) inside and downstream of a static turbine (at pipe mid-section); flow velocity = 3.5 m/s. TSR = tip
speed ratio.

Figure 8. Performance of three-, four- and five-bladed turbines versus tip speed ratio (TSR): (a) output power; (b) output torque during
complete turbine revolution at TSR = 3; (c) efficiency; (d) head losses.

number of possible blades was set to five. The increase in 65 N.m (for N = 5), and reduced the unwanted torque
the number of turbine blades increases turbine solidity, as fluctuations by generating more (five) peaks during a
dictated by Equation (2) and shown in Figure 8(a), which complete revolution of the turbine, at θ = 16°, 88°, 160°,
increases the capacity of the turbine to yield more power. 232° and 304°, as shown in Figure 8(b). The simulation
The turbine with five blades yielded a power of 1300 W, also showed that turbine power and efficiency increased
while that with three blades yielded only 870 W, at opti- with the increase in TSR, reaching a peak (at optimum
mum TSR. An increase in turbine solidity by 67% (from TSR of 3), and then decreased. This trend has also been
N = 3 to 5) increased the turbine power yield by 50% documented for lift-based turbines by other investiga-
(from 870 W to 1300 W). Jiyun et al. (2018) reported that tors (e.g. Jiyun et al., 2018). Maximum turbine efficiency
for lift-based cross-flow turbines, increasing the number (Figure 8c), at TSR = 3, was found to be almost the same
of blades would increase the turbine power. for N = 3, 4 and 5, with an average value of 29%. Thus,
The increase in the number of turbine blades also the increase in the number of blades from three to five
increased turbine torque, from 45 N.m (for N = 3) to did not have much effect on the overall efficiency of the
544 G. DIAB ET AL.

Figure 9. Flow field in and around the turbine (at mid-height) for three- and five-bladed turbines at different tip speed ratios (TSRs): (a)
pressure field (Pa); (b) velocity field (m/s).

turbine, as the increase in torque was balanced by the cor- water and thus would experience a high reduction in
responding increase in the head losses across the turbine efficiency. Talukdar et al. (2018) recommended that tur-
(Figure 8d). bine solidity must be in the range of 0.2–0.35 for the
Figure 9 shows the pressure and velocity distribution best operational conditions. On the other hand, Li and Li
at the mid-section of the three- and five-blade turbines (2010) suggested that a narrower range of helical turbine
at various TSR values. It can be seen from the figure that solidity (0.15–0.20) would be better, to ensure that the
for TSRs in the range of 2–4, the increase in the num- turbine efficiency would not drop significantly for high
ber of turbine blades from three to five decelerated the flow conditions.
flow velocity inside the domain enclosed by the turbine. For a chord length of 58 mm, the turbine solid-
Thus, the pressure would increase and consequently the ity was 22%, while it was 33% for a chord length of
lift on blade would also increase, enhancing the turbine 87 mm. Nonetheless, Figure 10(a) shows that the turbine
power. power was 700 W and 1300 W for blade chord lengths
of 58 mm and 87 mm, respectively. Hence, the turbine
power increased by 88.4% for a 50% bigger chord length
Chord length
(from 58 mm to 87 mm) at the optimal TSR value of 3.
To investigate the effect of the chord length of the turbine Other runs showed that a bigger chord length of 116 mm
blade on its performance, the chord length was changed resulted in a slight increase in turbine power, with new
from 58 mm (around 10% of turbine diameter) to 87 mm power of only 1380 W. The pressure drop across the tur-
(around 15% of turbine diameter). The turbine solidity bine varied linearly with TSR for simulated chord lengths,
represents resistance to flow in pipelines by blocking a since rotating the turbine with higher speed increases
portion of the pipe cross-sectional area. In the case of the interaction/blockage of flow through the turbine,
high turbine solidity, in addition to the rotational speed thus inducing more head loss (Figure 10b). High turbine
of the turbine, the turbine can behave as an obstacle to solidity is expected to have the same impact on pressure
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 545

Figure 10. Performance of five-bladed turbines with chord lengths of 58 and 87 mm versus tip speed ratio (TSR): (a) output power; (b)
head losses; (c) efficiency; (d) output torque during complete turbine revolution at TSR = 2, 3 and 4.

drop as high TSR owing to the additional flow block- Figure 11 shows the pressure and velocity distribu-
age. For a 50% increase in chord length (from 58 mm to tion at the mid-section of five-bladed turbines with chord
87 mm), the head losses varied by a similar percentage of lengths of 58 and 87 mm at different TSR values. It was
around 54% (from 0.59 m to 0.89 m) at a TSR of 3. Exam- observed that the increase in turbine solidity due to
ining the results of the turbine efficiency (Figure 10c), it increasing chord length had an impact on both the veloc-
can be seen that increasing the chord length (from 58 mm ity and pressure inside the turbine. Increasing the chord
to 87 mm) improved the overall turbine efficiency (from length induced a change in the flow field inside the tur-
20% to 27%) at a TSR of 3. A further increase in chord bine, especially around the blades, which was associated
length, to 116 mm, showed that the turbine efficiency with changes in the pressure magnitude. This resulted in
would only reach 28%, which can be attributed to the higher pressure variation across a single blade and higher
high head loss encountered for a chord length of 116 mm. lift force and power output.
Moreover, a bigger chord length of 87 mm reduced torque
fluctuations (Figure 10d).
Helicity
Both the number of blades and the chord length
are significant contributors to turbine solidity. However, The helical shape of the turbine blades is the main con-
practical limits restrict the number of helical blades to tribution made by Gorlov for enhancing turbine perfor-
five and it is recommended that solidity should not mance compared to straight-bladed turbines. The helical
exceed 0.35. In this study, a chord length of 20% of the angle, , is changed from zero (straight blades) to 60°
turbine diameter yielded the highest efficiency and lowest and 120°. Figure 12(a) shows that at a TSR of 3, all of
power fluctuations. Previous results show that the effect the rotors experienced the highest power of 1443, 1328
of blade chord length is more pronounced in enhancing and 1114 W for  = 0°, 60° and 120°, respectively. The
turbine performance than that associated with the num- turbine with straight blades,  = 0°, had the highest
ber of blades. A change in solidity by 67% (by increasing power output at TSR = 3, while blades with  = 60°
the number of blades from three to five) led to higher tur- and 120° had lower power output. Increasing the helical
bine output power, by 52%, at TSR = 3, while a change in angle from 0° to 120° resulted in a 22.8% reduction in tur-
solidity by only 50% (by increasing the chord length from bine power. However, when examining the performance
58 mm to 87 mm) increased the power output by 88.4%. of the rotors at any TSR, the rotor with straight blades has
546 G. DIAB ET AL.

Figure 11. Flow field in and around five-bladed turbines (at mid-height) for chord lengths of 58 and 87 mm at different tip speed ratios
(TSRs): (a) pressure field (Pa); (b) velocity field (m/s).

the lowest power compared to rotors with a helical blade highest power output, while increasing the blade overlap
shape. Thus, helical-bladed rotors have an advantage over reduced the power output.
straight-bladed ones by providing a steadier power over a Figure 12(b) shows the head loss across the rotor of
range of TSR values, while straight-bladed rotors provide the turbine for  = 0°, 60° and 120° at TSRs of 2, 3 and
the highest power. 4. When the helicity was reduced from 60° to 0°, the
Adjusting the helicity from 60° to 120° resulted in a head loss increased by 3.4%, 4.9% and 2.7% at TSRs of
reduction in the power, by 10.7%, 16% and 28%, at TSRs 2, 3 and 4, respectively. On the other hand, when helicity
of 2, 3 and 4, respectively. This means that increasing was adjusted from 60° to 120°, the head loss was reduced
the helicity beyond a certain value (60° in this study) almost by 12% for all TSR values. Thus, a high helical
would result in a reduction in the power produced by angle reduced the head loss across the turbine. In gen-
the turbine. The torque curve tended to be stable, with eral, the effect of blade helicity on head losses was less
minimum variation until it reached the least fluctuation pronounced than the effects on power output and overall
at  = 120° (Figure 12c and d). This is typical behavior efficiency.
for NACA hydrofoil blades, which achieve their hydrody-
namic maximum when flow is perpendicular to the blade
Pitch angle
section (in straight-bladed turbine). In helical-bladed
turbines, the flow over the blades is deformed in a span- The effect of blade curvature, or blade pitch angle β, on
wise direction, reducing blade section efficiency and thus the turbine performance was further studied by examin-
total turbine efficiency. Marsh et al. (2015) investigated ing the effect of β on turbine output power, torque and
the effect of helicity from 0° to 120° at a stream velocity efficiency. Pitch angles of 0°, 3°, −3° and −6° were con-
of 1.5 m/s and TSR in the range of 1.5–3.5, and observed sidered in the analysis. The maximum power (1350 W)
similar results, with the straight-bladed rotor having the and efficiency (27%) were achieved by the turbine when β
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 547

Figure 12. Performance of a five-bladed turbine with helicity of 0°, 60° and 120° versus tip speed ratio (TSR): (a) output power; (b) head
losses; output torque during complete turbine revolution at (c) TSR = 2 and (d) TSR = 4.

was 0°. At other pitch angles, the power and efficiency sig- region behind the shaft. In the shaft wake region, the flow
nificantly decreased, as shown in Figure 13(a). For pitch velocity was zero, with eddies and turbulence. This wake
angles of ± 3°, turbine power and efficiency decreased by region did not have any pronounced effect on the blades
38%; this is attributed to the early stall effect on the blade. downstream of the turbine at 180° ≤ θ ≤ 360°, except
On the other hand, the variation in blade pitch angle in the sections near the lower and upper hubs. Near the
was found to have a negligible effect on the head losses hubs, the wake generated by the shaft would result in a
across the turbine; a variation of head losses of 2% was reduction in the lift force in these blade sections. How-
observed at β = −6° compared to β = 0°. The highest ever, this had minimal or no effect on the total power
torque was produced by the turbine at zero pitch angle, as generated by the turbine. Figure 15 shows how the flow
shown in Figure 13(b). While both negative and positive separation around the shaft resulted in a zero-flow region
pitch angles produced lower torque values, negative pitch that extended to around 3.5ds from the shaft. After this
angles had higher torque than positive pitch angles. How- distance, the flow restored its velocity before interacting
ever, the pitch angle variation did not have a noticeable with the downstream blades again. Figure 15 shows also
effect on changing the torque fluctuations. that the minor shaft effect on flow velocity inside the tur-
bine was independent of the number of blades or solidity
of the turbine.
Hubs and shaft
The top and bottom hubs (with diameter of 250 mm)
The hubs and shaft are essential parts of the turbine for were expected to have more pronounced effects on the
fixation of the helical blades and connection with the turbine performance compared to the shaft owing to the
generator. Although they do not contribute to the gen- increase in the drag force (Figure 16). Average drag forces
eration of lift force, they may have an adverse impact on of 43 and 74.7 N were experienced by the turbine along
the hydrodynamic performance of the turbine. The shaft various positions during one complete revolution, with
diameter, ds , is taken as 52 mm in this study, which is variation up to 25.7% (difference between maximum and
10% of the turbine diameter. Examination of the two- minimum), at TSRs of 2 and 4, respectively. The top and
dimensional velocity field around the shaft in Figure 14 bottom hubs were expected to experience similar drag
at two sections showed that the effect of the shaft on flow forces owing to the symmetric nature of the hubs. Minor
velocity was minimal on the flow field inside the turbine. variations were observed between the drag force on the
Where the flow met the shaft, the streamlines split and lower and top hubs, up to 22.3% (Figure 16), at some
passed from both sides, leading to the creation of a wake azimuth positions. This was attributed to the effect of the
548 G. DIAB ET AL.

Figure 13. Performance of a five-bladed turbine with various pitch angles: (a) output power and efficiency; (b) output torque during
complete turbine revolution at tip speed ratio (TSR) = 3.

Figure 14. Velocity flow field (m/s) in and around the turbine: (a) near hub; (b) middle section, for a five-blade turbine at different tip
speed ratios (TSRs) and azimuth angles.

Figure 15. Normalized flow velocity at various sections downstream of the turbine shaft at the indicated locations.

blades on the velocity profile near the hubs and their heli- velocity. To illustrate this argument, the velocity distri-
cal shape. At a high TSR value of 4, the average drag force bution around the hub is plotted in Figure 17 at TSRs of
experienced by the hub was more than that at a TSR value 2, 3 and 4. The extent of disturbance in the velocity of
of 2. This was attributed to the increase in velocity around flow around the hub was higher in the case of a TSR of
the hub at a high TSR, which would increase the drag, 4 than for TSRs of 3 and 2. The velocity profile recovery
which is proportional to the second power (square) of occurred more rapidly in the case of a TSR of 2.
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 549

of 88.4% and 97.1%, respectively (with respect to


10% diameter chord length) at a TSR of 3.
(2) Adjusting turbine helicity from 0° to 60° and 120°
reduced the power by 8% and 22%, respectively
(compared to 0°). Introducing a pitch angle of 3°
to the blades reduced the overall efficiency by 38%
regardless of the direction.
(3) The shaft and hub generated drag force; however,
the impact of the drag forces on turbine perfor-
mance was minimal. Turbine stall created a wake
that extended to four times the turbine diameter,
Figure 16. Drag force at different azimuth positions on the shaft, beyond which the velocity profile was back to nor-
top hub and bottom hub at tip speed ratios (TSRs) of 2 and 4. mal at a velocity of 3.5 m/s.
(4) The pressure drop is one of the most important
metrics of in-pipe turbine performance. For all sim-
ulated cases, the head drop did not exceed 1.6 m,
which is equivalent to 2.7% of the total head defined
inside the pipe. Head drop varied linearly with TSR
and was proportional to change in solidity. On the
other hand, helicity and pitch angle had minimal
effects on head loss. Thus, installing in-pipe turbines
would have a minimal impact on flow in a water
supply network.

In conclusion, this paper presented an initial charac-


Figure 17. Velocity field (m/s) downstream of the top hub at
terization of the turbine performance for limited TSRs
different tip speed ratios (TSRs): (a) TSR = 2; (b) TSR = 3; (c)
TSR = 4. and flow velocities. Future research will consider a wider
range of TSRs and flow velocities for detailed turbine per-
formance curves. Moreover, this study suggests a distance
of four times the turbine diameter between installed tur-
Conclusions
bines in series. The behavior of multiple turbines installed
This research explored the potential of in-pipe turbines in series inside the pipe with different diameters will
to be installed in main pipelines. A 3D numerical model, be further studied with respect to cavitation, excessive
based on OpenFOAM, was utilized to investigate the pressure drop and any possible deterioration in turbine
performance of one of the available in-pipe water tur- performance.
bines, namely the lift-based helical-bladed turbine. The
performance of the turbine was assessed by examining Acknowledgments
the effects of its geometric properties, such as the num-
The writers are indebted to Eng. Lobus Pirkl from CFD sup-
ber of blades, chord length, helicity and pitch angle, at port, s.r.o. for his help in setting the numerical model sim-
different TSRs on the turbine’s power, torque, torque fluc- ulations. Research has been funded by the University of Abu
tuation, efficiency and head losses. The main findings are Dhabi, office of research.
as follows.
Disclosure statement
(1) Owing to the spherical shape of the turbine, geo-
No potential conflict of interest was reported by the authors.
metric constraints made it complicated to use more
than five blades; however, adjusting the solidity via
ORCID
the chord length was more feasible from a design
viewpoint. A five-bladed turbine produced more Ahmed M.A. Sattar http://orcid.org/0000-0002-2608-6979
power (1300 W) than a three-bladed (870 W) tur-
bine, by 50%, while all turbines (three-, four- and References
five-bladed) had the same efficiency (29%) at a TSR Alemi, H., Nourbakhsh, S. A., Raisee, M., & Najafi, A. F.
of 3. Chord length adjustment from 10% to 15% and (2015). Effects of volute curvature on performance of a low
20% of turbine diameters resulted in power increases specific-speed centrifugal pump at design and off-design
550 G. DIAB ET AL.

conditions. Journal of Turbomachinery, 137(4), 041009. turbines. Renewable Energy, 81, 926–935. https://doi.org/10.10
https://doi.org/10.1115/1.4028766 16/j.renene.2015.03.083
Chen, J., Yang, H. X., Liu, C. P., Lau, C. H., & Lo, M. OpenFOAM. (2019). www.openfoam.com.
(2013). A novel vertical axis water turbine for power Purdue ECT Team. (2016). LUCIDPIPETM POWER SYSTEM.
generation from water pipelines. Energy, 54, 184–193. ECT Fact Sheets. https://doi.org/10.5703/1288284316353.
https://doi.org/10.1016/j.energy.2013.01.064 Rezaeiha, A., Montazeri, H., & Blocken, B. (2018). Towards
Ghalandari, M., Bornassi, S., Shamshirband, S., Mosavi, A., optimal aerodynamic design of vertical axis wind turbines:
& Chau, K. W. (2019). Investigation of submerged struc- Impact of solidity and number of blades. Energy, 165,
tures’ flexibility on sloshing frequency using a boundary ele- 1129–1148. https://doi.org/10.1016/j.energy.2018.09.192
ment method and finite element analysis. Engineering Appli- Roy, S., & Saha, U. K. (2013). Review on the numerical inves-
cations of Computational Fluid Mechanics, 13(1), 519–528. tigations into the design and development of savonius wind
https://doi.org/10.1080/19942060.2019.1619197 rotors. Renewable and Sustainable Energy Reviews, 24, 73–83.
Gorlov, A. M. (1995). The helical turbine: A new idea for https://doi.org/10.1016/j.rser.2013.03.060
low-head hydro. Hydro Review, 14(5). https://www.osti.gov/ Salih, S. Q., Aldlemy, M. S., Rasani, M. R., Ariffin, A. K., Ya,
biblio/232050 T. M. Y. S. T., Al-Ansari, N., Yaseen, Z. M., & Chau, K.-
Gourdain, N. (2015). Prediction of the unsteady turbulent flow W. (2019). Thin and sharp edges bodies-fluid interaction
in an axial compressor stage. Part 1: Comparison of unsteady simulation using cut-cell immersed boundary method. Engi-
RANS and LES with experiments. Computers & Fluids, 106, neering Applications of Computational Fluid Mechanics, 13,
119–129. https://doi.org/10.1016/j.compfluid.2014.09.052 860–877. https://doi.org/10.1080/19942060.2019.1652209.
Hasanzadeh, N., Payambarpour, S. A., Najafi, A. F., & Mag- SALOME. (2019). www.salome-platform.org.
agnato, F. (2021). Investigation of in-pipe drag-based tur- Samora, I., Hasmatuchi, V., Münch-Alligné, C., Franca, M.
bine for distributed hydropower harvesting: Modeling and J., Schleiss, A. J., & Ramos, H. M. (2016). Experimental
optimization. Journal of Cleaner Production, 298, 126710. characterization of a five blade tubular propeller turbine
https://doi.org/10.1016/j.jclepro.2021.126710 for pipe inline installation. Renewable Energy, 95, 356–366.
Jiyun, D., Hongxing, Y., Zhicheng, S., & Xiaodong, G. (2018). https://doi.org/10.1016/j.renene.2016.04.023
Development of an inline vertical cross-flow turbine for Schlabach, R. A., Cosby, M. R., Kurth, E., Palley, I., & Smith,
hydropower harvesting in urban water supply pipes. Renew- G. (2011). In-pipe hydro-electric power system and turbine.
able Energy, 127, 386–397. https://doi.org/10.1016/j.renene. US7959411B2.
2018.04.070 Tahadjodi Langroudi, A., Zare Afifi, F., Heyrani Nobari, A., &
Li, S., & Li, Y. (2010, March 28–31). Numerical study on Najafi, A. F. (2020). Modeling and numerical investigation
the performance effect of solidity on the straight-bladed on multi-objective design improvement of a novel cross-flow
vertical axis wind turbine. 2010 Asia-Pacific Power and lift-based turbine for in-pipe hydro energy harvesting appli-
Energy Engineering Conference. Presented at the 2010 Asia- cations. Energy Conversion and Management, 203, 112233.
Pacific Power and Energy Engineering conference (pp. 1–4). https://doi.org/10.1016/j.enconman.2019.112233
https://doi.org/10.1109/APPEEC.2010.5449269. Talukdar, P. K., Kulkarni, V., & Saha, U. K. (2018). Field-testing
Ma, T., Yang, H., Guo, X., Lou, C., Shen, Z., Chen, J., & Du, J. of model helical-bladed hydrokinetic turbines for small-
(2018). Development of inline hydroelectric generation sys- scale power generation. Renewable Energy, 127, 158–167.
tem from municipal water pipelines. Energy, 144, 535–548. https://doi.org/10.1016/j.renene.2018.04.052
https://doi.org/10.1016/j.energy.2017.11.113 Yang, W., Hou, Y., Jia, H., Liu, B., & Xiao, R. (2019). Lift-type
Marsh, P., Ranmuthugala, D., Penesis, I., & Thomas, G. (2015). and drag-type hydro turbine with vertical axis for power
Numerical investigation of the influence of blade helic- generation from water pipelines. Energy, 188(C), 116070.
ity on the performance characteristics of vertical axis tidal https://doi.org/10.1016/j.energy.2019.116070

You might also like