Hydrofoil Impellers vs. Pitched Blade Turbines in Lignocellulosic Slurries

In a previous article (ref 1), it was stated that both pitched blade turbines and hydrofoils could be successfully used to agitate high-solids lignocellulosic materials, but that the hydrofoil could do the job with less power and torque. This article expands on that comment.

A little history

When I did my first tests on such slurries with turbines about a decade ago (having known for more than 2 decades that a helix would not work well in such materials), the only test impellers available were pitched blade turbines (figure 1). However, when I scaled up from these tests, I substituted hydrofoils (Figure 2), expecting about a 1/3 savings in torque, based on the known power draw and pumping characteristics of each impeller style. The resulting demo-scale equipment results have generally shown about a 50% reduction in torque compared to that expected from the pitched blade turbine scale-up. Many will be surprised to hear this, as the conventional wisdom is that hydrofoils are best used in turbulent flow, having little or no advantage in laminar flow. (However, the known impeller characteristics, such as power number and pumping number, still show a substantial calculated advantage in laminar flow). Since then, I have used hydrofoils in many lignocellulosic slurry installations, with solids as high as 23%, depending on the nature of the solids and any pretreatment used. All were successful.

Today’s situation

Though there is a substantial track record of success with hydrofoils, until recently, I had not done direct comparisons in the lab scale. Previous testing with hydrofoils had indicated that more torque is needed in the bottom of the vessel than in the upper zone, which, in practice, means the bottom impeller must be larger than the upper impellers (2).

Recently conducted testing in a “mixed” slurry (containing wood flour, sawdust, milled untreated switchgrass, pretreated woody biomass and pretreated sugar cane bagasse) resulted in the observation that, at maximum % solids possible to agitate well, the pitched blade turbine took more than 3 times as much torque as the hydrofoil to agitate the bottom of the tank, and twice as much to agitate the top zone of the tank. The comparative advantage of the hydrofoil actually increased as the mixing difficulty increased, i.e., as the viscosity got higher. At equal diameter, the pitched blade turbine actually had to operate at a higher shaft speed than the hydrofoil to create complete motion in the tank bottom.

At lower viscosities, the advantage of the hydrofoil was more within expectations: about ½ of the torque of a pitched-blade turbine. I believe the reason for such differences lies within the flow pattern created by the different impeller types. The pitched blade turbine creates more radial flow and tends to recirculate within its zone of influence. The hydrofoil maintains a more axial pattern and projects farther in both radial and axial directions. With a two-impeller system, the pitched blade turbine creates a staged flow pattern, whereas the hydrofoil creates a single-zone pattern from top to bottom, in a 1:1 batch aspect ratio gemoetry.

A production simulation

Based on the above testing, plus many specific tests on actual customer samples, I have created a simulated batch hydrolysis reaction profile for the first 10 hours after enzyme addition. Though lignocellulosic slurries are best characterized as Herschel-Bulkley fluids (pseudoplastic with yield stress), for convenience, I will just use the pseudoplastic model here:

1) μ_a = M(dv/dx)^(n-1)

where M is the viscosity at 1/s and n is the power law exponent.

Table 1 shows a profile of rheology changes as the reaction progresses. This profile is based on averaging a number of actual materials I have tested, but does not represent any one actual material.

As the reaction progresses and the material thins, the required speed and power draw will reduce. The speed profile is also based on an amalgamation of tests in different slurries. Table 1 is based on a vessel of 248” diameter, with a working volume of 50,000 gallons.

Table 1 Power Draw Profile

The table shows both comparative power and comparative torque. Power relates to operating cost. Torque relates to capital cost of the agitator. The power draw comparison can be shown graphically, as in figure 3.

Figure 3 Power requirement profile

In this simulation, the hydrofoil saves about 60% of the energy required in the first 10 hours, then about 20% after that. The capital cost of a machine designed using a pitched blade turbine would be about twice as much as one designed using a hydrofoil, due to the difference in torque.

Although the actual numbers may be different in specific lignocellulosic slurries, there is no doubt that use of hydrofoil impellers will save on capital and operating costs compared to using pitched blade turbines. Existing facilities could reduce their power costs by retrofitting hydrofoils if they presently have pitched blade turbines.

About the author: Gregory T. Benz, P.E., is President of Benz Technology International, Inc. and a member of Lee Enterprises Consulting, the world’s premier bioeconomy consulting group with over 100 consultants worldwide. The group has expertise in many areas, including the subject discussed in this report.

Acknowledgements: Figure 1 courtesy of Chemineer, a brand of NOV. Figure 2 courtesy of Fusion Mixers, Inc.

References cited

1) “Impeller Selection for Lignocellulosic Hydrolysis Reactors”, G.Benz, Biofuels Digest Online, 3/16/2017

2) “Determining Torque Split for Multiple Impellers in Slurry Mixing”, G. Benz, Chemical Engineering Progress, February, 2012, pp 45-48