This project WiCOM has been developed in Schlumberger in 2005. Its purpose was the telemetry between downhole receiver and the surface of the land in two senses: Uplink and Downlink. In Uplink we transmit the temperature, pressure and the voltage measurements.
In Downlink there are two possibilities of using: activation of some mode or receiver and activation of some explosion to ensure the delivery of oil from some side pocket.
In any way the two applications are supposed of such the Tx Rx link: either the oil/gaz investigation telemetry of the oil/gaz direct deliberation.
The clients of such the telemetry are the oil production companies like: Total, Shell, Gazprom, British Petrolium.
The multidisciplinary team was composed by mechanical engineers, theoretical physicists, electronical engineers, embedded software engineers and signal processing engineers as well.
Technical realization of project.
This project from the point of view of technical realization was one of the most complex in my experience. The first purpose was to transmit 30..50 bits from the surface of the lend to downhole receiver, situated at 3..5 km under the surface. There the temperature sometimes reaches 150°C.
The transmission has been ensured with electromagnetic waves, that propagated vie the different layers of the earth. The frequency should be the closest to some Hertz, because according to the physical lows of propagation in solid states the worse the higher propagation frequency.
Concerning the high temperature, the only available DSP chip able to work in such the environment was DSPic of MicroChip.
Another constraint was the limited size of DSP as well as limited RAM, MIPS and stack.
Interesting moment was the transmission of information: normally to verify the BER 10-2 one needs to send 10 in power 4 bits (10.000). When the carrier frequency is around 1 Hz, it needs 10.000 secs = 10.000/60/60 hours = 3 hours. But we tested the packets, so for 30 bits we needed 60 bits minim to guarantee the redundancy. With synchronization part it went to 100 bits. So, overall experiment went to 300 hours instead of 3 hours, the equivalent of 12 days )) Finally we had only the time to test during 1..2 hours.
My participation in project.
My role was to analyse with modem parts we need to implement and to determine the concrete algorithm and implementation for every element. Of course we had big multidisciplinary team and we had another signals processing team in UK for consultations.
So the activities where I participated were:
- null
- Choice of modulation (it was Manchester code without carrier)
- Choice of channel coding and CRC
- Choice of synchronization preamble
- Implementation and choice of AGC and synchronization algorithm
- Simulation of transmitter – receiver chain in Matlab
- Implementation in C under DSP MicroChip, effective DSP implementation (cycles unrollment, escaping unnecessary redundant operations)
- Validation on laboratory test and in field test with real downhole receiver
Environment and techniques used in project.
- Techniques: MSK, BPSK, Manchester, phase/frame/frequency synchronization, phase tracking, code corrector (BCH, Convolution codes), AGC, filtering (FIR/IIR), Costas, Goertzel, CRC
- Environment: C/C++, VisualC++, Matlab, Asm, DSPic Microchip, downhole receiver, oil well
Main results.
All the packets send to receiver have been successfully received.
