Research Papers

Low-Flow-Coefficient Centrifugal Compressor Design for Supercritical CO2

[+] Author and Article Information
C. Lettieri

MIT Gas Turbine Laboratory,
Massachusetts Institute of Technology,
Cambridge, MA 02139
e-mail: lettieri@mit.edu

N. Baltadjiev

MIT Gas Turbine Laboratory,
Massachusetts Institute of Technology,
Cambridge, MA 02139
e-mail: nikola@mit.edu

M. Casey

University of Stuttgart,
Stuttgart 70174, Germany
e-mail: casey@itsm.uni-stuttgart.de

Z. Spakovszky

MIT Gas Turbine Laboratory,
Massachusetts Institute of Technology,
Cambridge, MA 02139
e-mail: zolti@mit.edu

Careful analysis of the flow field shows that three-dimensional effects in the relatively narrow impeller passage are negligible.

Full stage CFD computations conducted after this parametric impeller study indicate that the trend in overall vaned diffuser loss is well captured using the scaling law for profile loss.

The crossover bend and return channel were not included in this assessment as the focus here is on the stability of the impeller and the vaned diffuser.

Contributed by the International Gas Turbine Institute (IGTI) of ASME for publication in the JOURNAL OF TURBOMACHINERY. Manuscript received October 14, 2013; final manuscript received November 5, 2013; published online January 31, 2014. Editor: Ronald Bunker.

J. Turbomach 136(8), 081008 (Jan 31, 2014) (9 pages) Paper No: TURBO-13-1235; doi: 10.1115/1.4026322 History: Received October 14, 2013; Revised November 05, 2013

This paper presents a design strategy for very low flow coefficient multistage compressors operating with supercritical CO2 for carbon capture and sequestration (CCS) and enhanced oil recovery (EOR). At flow coefficients less than 0.01, the stage efficiency is much reduced due to dissipation in the gas-path and more prominent leakage and windage losses. Instead of using a vaneless diffuser as is standard design practice in such applications, the current design employs a vaned diffuser to decrease the meridional velocity and to widen the gas path. The aim is to achieve a step change in performance. The impeller exit width is increased in a systematic parameter study to explore the limitations of this design strategy and to define the upper limit in efficiency gain. The design strategy is applied to a full-scale reinjection compressor currently in service. Three-dimensional, steady, supercritical CO2 computational fluid dynamics (CFD) simulations of the full stage with leakage flows are carried out with the National Institute of Standards and Technology (NIST) real gas model. The design study suggests that a nondimensional impeller exit width parameter b2* = (b2/R)ϕ of six yields a 3.5 point increase in adiabatic efficiency relative to that of a conventional compressor design with vaneless diffuser. Furthermore, it is shown that in such stages the vaned diffuser limits the overall stability and that the onset of rotating stall is likely caused by vortex shedding near the diffuser leading edge. The inverse of the nondimensional impeller exit width parameter b2* can be interpreted as the Rossby number. The investigation shows that, for very low flow coefficient designs, the Coriolis accelerations dominate the relative flow accelerations, which leads to inverted swirl angle distributions at impeller exit. Combined with the two-orders-of-magnitude higher Reynolds number for supercritical CO2, the leading edge vortex shedding occurs at lower flow coefficients than in air suggesting an improved stall margin.

Copyright © 2014 by ASME
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Fig. 1

Pitchwise averaged entropy change in meridional plane: seal leakage mixing losses and disk friction losses account for nearly half of the overall loss

Grahic Jump Location
Fig. 2

Conventional gas path design with pinched vaneless diffuser (left) versus new design concept with wider gas path and vaned diffuser (right)

Grahic Jump Location
Fig. 3

Effect of b2* on mass averaged diffuser inlet flow angle and total-to-total polytropic impeller efficiency

Grahic Jump Location
Fig. 4

Effect of b2* on computed vaneless space and estimated vaned diffuser loss coefficients

Grahic Jump Location
Fig. 5

Assessment of computed datum stage pressure coefficient with experimental data

Grahic Jump Location
Fig. 6

Assessment of computed datum stage total-to-total polytropic efficiency with experimental data

Grahic Jump Location
Fig. 7

Computed stage pressure coefficient without leakage flow: wider impeller with vaned diffuser yields 0.02 increase in pressure rise

Grahic Jump Location
Fig. 8

Computed total-to-total polytropic stage efficiency without leakage flow: wider impeller with vaned diffuser yields a 4.5 percentage point increase in efficiency

Grahic Jump Location
Fig. 9

Absolute Mach number contours at midspan: datum design (top) and redesign (bottom)

Grahic Jump Location
Fig. 10

Cumulative change in total-to-total polytropic efficiency relative to the datum performance

Grahic Jump Location
Fig. 11

Impeller and vaned diffuser pressure rise characteristics: diffuser dynamically unstable—error bars indicate fluctuations in pressure rise

Grahic Jump Location
Fig. 12

Mechanism for short-wavelength rotating stall inception in vaned diffusers (adopted from Ref. [11])

Grahic Jump Location
Fig. 13

Spanwise distribution of pitchwise averaged flow angle: typical traditional centrifugal compressor (Ro ∼ 0.5) versus low flow-coefficient compressor design (Ro ∼ 0.07)

Grahic Jump Location
Fig. 14

Ekman layers for low flow-coefficient compressor design. Reduced end wall flow angles for supercritical CO2 suggest increased stability margin relative to low-pressure air.

Grahic Jump Location
Fig. 15

Diffuser pressure rise characteristics for air and supercritical CO2: leading edge separation and vortex shedding observed at flow coefficient (a) for air and (c) for supercritical CO2

Grahic Jump Location
Fig. 16

Mach number contours at 10% span, (a) air at ϕ/ϕd = 0.81, (b) supercritical CO2 at ϕ/ϕd = 0.81, and (c) supercritical CO2 at ϕ/ϕd = 0.68

Grahic Jump Location
Fig. 17

Spanwise distribution of pitchwise averaged flow angle for air and supercritical CO2 at onset of leading edge vortex shedding




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In