TY - JOUR

T1 - Electric submersible pump behavior for pumping non-Newtonian fluids

AU - Langbauer, Clemens

AU - Pratscher, Hans-Peter

AU - Ciufu, Alexandru-Cosmin

AU - Hoy, Michaela Gertrude

AU - Marschall, Christoph

AU - Pongratz, Reinhard

PY - 2020/12

Y1 - 2020/12

N2 - The performance of two electric submersible pumps (ESP), when pumping non-Newtonian polymer fluids, was tested at a pump test facility of the university. The results are compared with existing models, and the de-rating coefficients are evaluated for the operating range of the pumps that allow a proper design of ESP field installations. Two different electric submersible pump types (82 stages and 7 stages) were tested at the newly-built pump testing facility. The initial tests were performed with water at several frequencies to check the pump's performances, and the results were compared with the manufacturer's catalog curves. The performance measurements included pressure, flow rate, torque, rotational speed, and temperature. Besides, an intake pressure sensitivity analysis was performed. The actual polymer tests were conducted for three HPAM polymer concentrations (400-ppm, 800-ppm, 1600-ppm), three rotational speeds, and several discharge pressures. Additionally, fluid samples were taken during all tests and were subsequently analyzed. The polymer test results showed that the performance reduction factors for both pump types and all frequencies are much higher than expected. However, the provided polymer concentrations have low viscosity. Most likely, non-newtonian fluid behavior, specifically its viscoelasticity, is responsible for this substantial performance loss. Head and rate reduction close to 50 percent was measured at the best efficiency point of the pumps. A comparison to existing models indicates a significant underestimation of the performance loss when pumping a polymer solution. Moreover, the mechanical shearing of the molecular polymer chains was minor. Known models for pump performance do not apply in this instance. Because of the complexity of the polymer-laden fluid, we cannot offer one that works, and so we recommend that anyone in this situation must perform an experimental performance test. The novelty of those experiments was the testing of two powerful electric submersible pumps, using a non-newtonian polymer fluid, to investigate the pump performance, performance correction coefficients, and compares appropriate correction models to apply the gained knowledge more quickly in the field.

AB - The performance of two electric submersible pumps (ESP), when pumping non-Newtonian polymer fluids, was tested at a pump test facility of the university. The results are compared with existing models, and the de-rating coefficients are evaluated for the operating range of the pumps that allow a proper design of ESP field installations. Two different electric submersible pump types (82 stages and 7 stages) were tested at the newly-built pump testing facility. The initial tests were performed with water at several frequencies to check the pump's performances, and the results were compared with the manufacturer's catalog curves. The performance measurements included pressure, flow rate, torque, rotational speed, and temperature. Besides, an intake pressure sensitivity analysis was performed. The actual polymer tests were conducted for three HPAM polymer concentrations (400-ppm, 800-ppm, 1600-ppm), three rotational speeds, and several discharge pressures. Additionally, fluid samples were taken during all tests and were subsequently analyzed. The polymer test results showed that the performance reduction factors for both pump types and all frequencies are much higher than expected. However, the provided polymer concentrations have low viscosity. Most likely, non-newtonian fluid behavior, specifically its viscoelasticity, is responsible for this substantial performance loss. Head and rate reduction close to 50 percent was measured at the best efficiency point of the pumps. A comparison to existing models indicates a significant underestimation of the performance loss when pumping a polymer solution. Moreover, the mechanical shearing of the molecular polymer chains was minor. Known models for pump performance do not apply in this instance. Because of the complexity of the polymer-laden fluid, we cannot offer one that works, and so we recommend that anyone in this situation must perform an experimental performance test. The novelty of those experiments was the testing of two powerful electric submersible pumps, using a non-newtonian polymer fluid, to investigate the pump performance, performance correction coefficients, and compares appropriate correction models to apply the gained knowledge more quickly in the field.

KW - Chemical enhanced oil recovery

KW - Electric submersible pump

KW - Montanuniversitaet Leoben

KW - Non-Newtonian fluid

KW - Polymer-laden fluid production

KW - Pump test facility

UR - http://www.scopus.com/inward/record.url?scp=85091224387&partnerID=8YFLogxK

U2 - 10.1016/j.petrol.2020.107910

DO - 10.1016/j.petrol.2020.107910

M3 - Article

VL - 195

JO - Journal of Petroleum Science and Engineering

JF - Journal of Petroleum Science and Engineering

M1 - 107910

ER -