Design optimization to improve axial turbine stage efficiency by 6%

One of the leading manufacturers of turbodrills for the oil and gas drilling industry asked ADT to carry out the complete redesign of their existing hydraulic drilling turbine to suit different types of rock formations with a single design.

Turbodrills, multistage axial turbines with rotor and stator configuration that can have as many as 150 stages, are used for specific applications in oil and gas drilling. They use the fluid required to clean the borehole to drive a turbine to turn the drill bit. Conventional turbines run at relatively high speeds / low torques, which limits their applications to specific rock types. The performance targets set by the customer for this project requested an increase in torque at the low operating speeds and a low runaway (zero-torque) speed to ensure the safety of the drill-line and surrounding components.

The overall sizing of the turbine being fixed by the requirement within the borehole sizes. ADT investigated various possible meridional, including repeating mixed-flow designs to evaluate the potential for achieving high torque with low runaway speeds but unfortunately, the results showed that any conventional design could not meet the target requirements and novel avenues that needed to be explored.

ADT used the 3D inverse design code TURBOdesign1 to design a number of rotor/ stator configurations with varying design flow, rpm and meridional channels to understand the design parameters that most influenced the slope of the torque characteristic curve for oil and gas drilling. 

There was very careful control of the blade loading using TURBOdesign1's inverse design method that enabled the flow to remain stable at low rpm. This produced high torque while inducing a controlled separation at high rpm that reduced the torque obtained without any other adverse impacts to the turbodrill, such as vibrations or blade erosion. 

The final design provided +30% more torque at low rpm, whilst simultaneously reducing runaway speed by -10% compared to the traditional design. 

The final phase of this design project included using automatic optimization to improve the stage efficiency. The blade geometry used in TURBOdesign1 is parameterized by using the blade loading (as shown in Fig. 2) and TURBOdesign Optima is used to run Multi Objective Genetic Algorithm (MOGA) to explore the design space and create a Pareto front of contrasting design objectives.

 

 

TURBOdesign1 parametrizes a fully 3D blade shape with only a few design parameters (Fig. 2) and also provides detailed 3D inviscid surface pressure and velocity information, using ADT’s correlations based on the 3D inviscid flow fields it is possible to run hundreds of designs in less than one hour and achieve a well-defined Pareto Front (Fig. 3) clearly showing the trade-off between loss mechanisms.