Neural network emulating a g-function

A neural network model to analyze interupted TRTs

TRT Analysis uses an artificial neural network (ANN) trained to reproduce accurately the output of a transient 3D finite element model. The resulting ANN model integrates the pipe and borehole geometry, the thermal properties (λ and α) of the fluid, pipes, grout and geological material, as well as the vertical fluid velocity in the pipe loop. 

As the ANN model is coupled with an efficient temporal superposition scheme, the temperature changes due to heating power variations can be modelled efficently for a TRT that is less than 21 days. 

For a given set of input parameters, the ANN model allows a quick simulation of the fluid temperature at the borehole inlet and outlet with an unprecedent accuracy. This allows a simultaneous analysis of the heating and recovery phases, of pulsated TRTs, or of interupted TRTs.

Back to blog