Professional Documents
Culture Documents
Kor Et Al-2008-Asia-Pacific Journal of Chemical Engineering
Kor Et Al-2008-Asia-Pacific Journal of Chemical Engineering
Research Article
Using CFD to identify means of reducing power
consumption for mixing and suspension in paper pulp stock
chests
Yann Kae Kor,1 Rolf G. H. Prince1,2 and David F. Fletcher1 *
1
School of Chemical and Biomolecular Engineering, University of Sydney, NSW 2006, Australia
2
Visy Pulp & Paper, Smithfield, NSW 2164, Australia
ABSTRACT: This paper presents ongoing work at Australia’s largest paper recycler, Visy Pulp & Paper, involving
experimental and computational fluid dynamics (CFD) studies to understand paper pulp mixing processes, aimed at
identifying means of reducing the large power consumption of the agitators currently used. The key focus of this paper
is on the modelling of a pseudo-plastic fluid and the determination of means by which adequate flow circulation can be
maintained in such a system. We look at reducing the present agitator speed, at using a larger impeller, and examine
the performance of a completely different design. Power consumption reductions of 50%, and perhaps 75–80%, from
present levels, appear feasible and could be implemented economically. 2008 Curtin University of Technology and
John Wiley & Sons, Ltd.
KEYWORDS: recycled paper pulp stock; mixing; suspension; computational fluid dynamics; Bingham fluid; impeller
design; power consumption
Figure 4. The unstructured mesh on the existing and ITT Flygt impellers.
Note the fine meshing to capture details of the flow near the blade tips.
The impellers are not shown with the same scale. This figure is available in
colour online at www.apjChemEng.com.
The impeller rotation was modelled using the well- power rating of 30 kW), then half, and finally a quar-
known multiple reference frame (MFR) approach, in ter of that. Since the existing impeller and the larger
which a cylindrical zone around the impeller is simu- impeller are both marine type propellers, they have the
lated in a frame of reference moving with the impeller same power number (i.e. NP = 0.28). Hence when sim-
and the remainder of the fluid is in a stationary frame. ulations were performed for the larger impeller (diam-
The frozen rotor approach, in which the frame transfor- eter 2.4 m), the speed that corresponds to a power
mation is made assuming a single fixed position of the consumption of 30 kW and NP = 0.28 was first inves-
rotor blade, was used. The known inflow of 50 kg s−1 tigated (calculated to be 62 rpm). Subsequent simula-
of stock and the outflow to the pump were modelled via tions were performed for half and a quarter of that
an inlet and outlet, respectively. The free surface was speed. The new model ITT Flygt impeller (diame-
assumed to be flat and was represented via a no stress ter 0.58 m) is not a marine type propeller, and does
boundary. not have the same power number as the existing
The high-resolution scheme, which provides bounded and larger impellers. Hence, the manufacturer’s recom-
values as close as possible to second order accuracy, mended speed of 633 rpm was chosen, and subsequent
was used for the convective scheme. In all reported simulations were performed for two-third and one-third
cases the normalised residuals were below 10−4 and of that speed.
mass and momentum balances were below 1%. The
torque on the blade was monitored to ensure that it
had stabilised.
The existing impeller
Figure 6. Steamlines for the existing impeller for various rotation speeds. The rotation rate,
power consumption and critical Reynolds number are also given. This figure is available in
colour online at www.apjChemEng.com.
2008 Curtin University of Technology and John Wiley & Sons, Ltd. Asia-Pac. J. Chem. Eng. 2008; 3: 144–150
DOI: 10.1002/apj
Asia-Pacific Journal of Chemical Engineering CFD MODELLING OF PAPER PULP CHESTS 149
Figure 7. Steamlines for the large impeller for various rotation speeds. The rotation rate,
power consumption and critical Reynolds number are also given. This figure is available in
colour online at www.apjChemEng.com.
Figure 8. Steamlines for the ITT Flygt impeller for various rotation speeds. The rotation rate,
power consumption and critical Reynolds number are also given. This figure is available in
colour online at www.apjChemEng.com.
velocities towards the inclined bottom. The remainder may be achievable at reduced power consumption.
would be propelled at higher velocities towards the near These simulations indicated that 50%, and perhaps
end of the bottom before discharging. The velocity con- 75–80%, saving from present levels of energy con-
tours demonstrated that a larger proportion of stock has sumption may be possible. Three possibilities to achieve
a speed greater than 0.1 m s−1 . In particular for the energy savings have been examined here. The first is to
633 rpm case, stock propelled into the middle of the operate the present agitator at half the existing speed,
stock chest reaches a minimum speed of 1 m s−1 . Most lowering it from 211 to 106 rpm. This would reduce
of the contents are in motion, with stock travelling at power consumption from 30 to 5 kW, an 83% reduc-
velocities of at least 0.2 m s−1 . tion in electricity expenditure. The second alternative is
The critical Reynolds number is satisfied at all speeds. to replace the existing impeller with the same model but
Hence, stock agitated with the ITT Flygt impeller, at having twice the diameter, and to run it at a low speed.
all proposed speeds, would be in good suspension and When set to run at 31 rpm, the power consumption is
adequately mixed before being discharged. Even at the predicted to be reduced from 30 to 6 kW, which implies
lowest speed, the degree of suspension and mixing an 80% reduction in electricity expenditure. However, it
appears to be sufficient. This speed consumes only is worth noting that running a large impeller at very low
7 kW of electrical power, and hence can be considered speeds is similar to providing a ‘paddle’ or ‘sweeping’
further as an alternative to the existing impeller. action to the contents of the stock chest. This action may
offer satisfactory macro-scale motion, but perhaps not
sufficient mixing. The third alternative is to replace the
CONCLUSIONS AND RECOMMENDATIONS existing impeller with an ITT Flygt impeller. If operated
at 211 rpm, the power consumption could be reduced
CFD simulations have been performed that allow the from 30 to 7 kW, offering a 77% reduction in electricity
flow-field in the stock chest and the power consump- expenditures.
tion to be calculated. Examination of the CFD results On the basis of the results reported here, further
demonstrates that the mixing of the paper pulp stock investigations were undertaken by Visy Pulp & Paper to
2008 Curtin University of Technology and John Wiley & Sons, Ltd. Asia-Pac. J. Chem. Eng. 2008; 3: 144–150
DOI: 10.1002/apj
150 Y. K. KOR, R. G. H. PRINCE AND D. F. FLETCHER Asia-Pacific Journal of Chemical Engineering
2008 Curtin University of Technology and John Wiley & Sons, Ltd. Asia-Pac. J. Chem. Eng. 2008; 3: 144–150
DOI: 10.1002/apj