Measurement while drilling (MWD) enables real-time measurement of downhole conditions for directional drilling, but most commercial MWD telemetry techniques such as mud pulse or electromagnetic methods suffer from limited transmission speeds. Acoustic telemetry has the potential for significantly faster transmission speeds, albeit with limited range due to drill string attenuation and noise. A common solution to this is to use acoustic repeaters, which incur high costs and require complex implementation. Instead of using repeaters, we utilized two carrier frequencies at the modes to transmit redundant data in combination with a lock-in amplifier (LIA) to extract the signals from the carriers. The extracted signals were then fused at the receiver to increase signal fidelity. An experimental setup was developed to transmit acoustic signals through a simulated drill string. The signals were first attenuated by the rubber section of the simulated drill string. The results show that the proposed system was able to achieve error-free transmission of packets at 64 bps up to 1.95 km without the use of a repeater which is an order of magnitude faster than current commercial MWD methods.

Moreover, the acoustic telemetry system requires the identification of the carrier frequencies near the natural frequencies of drill string with specified boundary conditions. This work proposes a finite element (FE) model based on the Timoshenko beam theory that predicts the dynamics of an actual drill-string over a wide frequency range. The frequency response of the model is compared to models in literature with similar components. Then, three configurations that follow a specified trajectory are defined with increasing lengths and curvature to represent the drill string assembly as it approaches the target reservoir. The frequency responses of the of the three configurations are determined and a carrier frequency was selected at the center of the third passband. Like the lab-scale experiments, packets of bits are first generated as telemetry data and then convolutionally encoded to reduce errors at the receiver. The signal is modulated using differential binary phase shift keying (DBPSK) and upconverted to the carrier frequency which is used as the force input to the model. Finally, the receiver at the surface demodulates and decodes the received acceleration to recover the transmitted bits using a digitally implemented LIA. The transmitted and received bits are again compared to calculate the bit-error rate (BER) for each signal-to-noise ratio (SNR) condition and used as the measure of performance. To simulate transmission, the time impulse responses are first recorded for the three different drill string configurations. These are then used to develop a finite impulse response filter (FIR) for simulation of the acoustic transmissions. The results show that the passband locations stay at the same frequencies and that transmission speed is limited by the passband widths.