We take easy communications for granted these days. It’s no bother to turn on a lightbulb remotely via a radio link or sense the water level in your petunias, but how does a drilling rig sense data from the drill head whilst deep underground, below the sea bed? The answer is with mud pulse telemetry, about which a group of researchers have produced a study, specifically about modelling the signal impairments and strategies for maintaining the data rate and improving the signal quality.
If you’re still confused, mud pulse telemetry (MPT) works by sending a modulated pressure wave vertically through the column of mud inside the drilling tube. It’s essential to obtain real-time data during drilling operations on the exact angle and direction the drill bit is pointing (so it can be corrected) and details of geological formations so decisions can be made promptly. The goal is to reduce drilling time and, therefore, costs and minimize environmental impact — although some would strongly argue about that last point.
One challenge with MPT is that the transmission media can be inconsistent. It may contain rocks and gas, leading to variations in physical properties like density, compressibility, and viscosity throughout the column, which can affect signal transmissibility. The MPT system includes a pressure transducer at the drill head that encodes data about local parameters such as temperature and pressure. The paper also describes other sources of noise that can distort the signal, including vibrations from the drill head and pressure pulses from the drilling mud pumps.
From what we can gather the MPT system is bidirectional, using mechanical means via ‘poppet valves’ to create positive or negative pressure pulses and a rotating slotted disk to generate continuous waves. It’s possible to achieve a data rate of 20 bps from depths of over 6 km. The paper also discusses other data transmission methods as part of the logging-while-drilling (LWD) system, some of which are used alongside MPT in specific circumstances. All of these methods face challenges when transmitting data through this complex medium.
We couldn’t find much on Hackaday about this topic, but we did recall an interesting piece about dealing with oil spills and who could forget this one about fracking?
Thanks to [Derek] for the tip!
The featured image is courtesy of Dynamic Graphics, Inc.
Another one in the series of “never heard of it before”. Amazing.
Mud-pulse is ubiquitous across the O&G drilling industry. I cannot express how hilarious it is to have a single-bit/second data transfer rates with 2024 tech, but that is the reality on some nightmare wells.
Every drilling service company I have worked at was working on some version of a faster mud siren; Physics makes this real difficult. If someone came up with a 1 kb/sec mud pulser they’d be able to print money.
Another interesting bit of this tech is the MWD tools strive to get the sensors as close to the drill bit as possible. Some of the designs are working to push the temperature ratings of equipment to 200*C, including the mud-pulsers. The batteries for these devices are designed for extreme heat and don’t work at room temperature.
What could be better than trying for acoustic signal transfer through a non-linear non-Newtonian fluid at high frequencies? So many PhD students are going to be casualties in this one….
This has been standard practice in the drilling industry since the mid 1980’s. One difference, the data rate quoted is about 30 times faster than what was available 40 years ago.
For those wanting more details, the development was done by BBN, for the geologic survey company Schlumberger. They would send instruments down the well to characterize the rock. Prior to this, it meant pulling the drill string out of the well then sending down the measuring head on a cable.
BBN — the same folks who analysed the Nixon tapes and developed Arpanet? They *do* get around!
(and Schlumberger used to own Heathkit)
Can you provide a reference about the BBN development? I know that SLB had their own development under Dennis Tanguy from about 1968. They ended up licencing the siren patent from Mobil Oil in 1978. Teleco went commercial in 1978. Schlumberger didn’t go commercial until Feb 1980.
@Clara – If people would stop thinking about oil as fossil “fuel” we would be much better off. We absolutely need oil exploration for plastics, pharmaceuticals, and chemicals of which oil is a precursor. I can envision a future without ICE engines. I cannot envision one without all of the latter.
“absolutely need oil exploration for plastics”
He says as microplastics course though his blood stream.
Yes. We should all stop using everything made of plastics NOW, starting with your keyboard…
The novelization of the documentary movie Alien remarked that this is why we need refineries that can travel interstellar space, and state of the art transport tugs like the movie subject Nostromo to ferry them around.
By the way according to that same film they are doing fantastic stuff with AI in the space industry. It will be a great advance when they get that tech down to Earth for everyday use.
When approx 80% of a barrel of crude oil is refined for use as various types of fuel (gasoline, Diesel, AVGas etc) it’s not unreasonable at all to think of oil as fossil fuel.
That approx 20% which is left over is essentially a byproduct of the fuel refining process that we have found ways to use instead of just throwing it away or burning it (more fuel)
Pretty much all of things made from that byproduct can be synthesized in other ways (maybe not as efficiently or cheaply) so oil isn’t something we ‘absolutely need’, it’s something we currently have and make use of.
Bidirectional MPT is not really used, the downlinking is usually done by sequencing the drill string rotation. Also 20bps that works in the field is rather unheard of. MPT using classic pulser (on/off opening and closing a mud channel) is 2bps, using a siren (rotating disc that opens and closes mud channels, data is sent by modulating the angular speed) you can get to 12bps. Higher speed sirens are not reliable at longer distances.
CW vs FM…interesting. The “ether” is mud!
Thanks for explaining the “siren” which was mentioned above but not explained.
We did very low bitrate one way comms using the string rotation sequencing trick. Essentially like a secret knock on a door. The drill head was listening for a series of lurches with very specific pauses between them to trigger an action. In this case it was to release a set of folding augers that popped out for bore widening.
oh god the nightmares of power management issues resulting in accidentally activating the reamer.
That this works at all is incredible, it amazes me what they can do at the bottom of a hole.
Fascinating and an incredibly mechanical solution.
I’d be amazed if it’s not been investigated and tried already but 6KM of drill string would seem to be a fairly ideal candidate for a VLF antenna, maybe it’s a privacy issue?
There is EM telemetry for downhole use, the carrier is a few Hz. It is a little bit faster than mud pulse, but we are still talking about bits per second. The issue here is that it cannot be used in every soil type (conductivity matters). Mud pulse telemetry works in every well. The only way to get decent data rate this day is to use a wired pipe (57kbps), but it is expensive and problematic. As someone above said, invent a downhole telemetry system that can reliably do > 100bps (don’t even need 1kbps) and you are a rich man.
I guess it’s a little bit too “grounded” along it’s way.
In 2013, I worked for a startup company that was sending both power down and receiving data up the drill pipe (no wires). Worked great and data rates were “decent”. We manufactured a tool that sensed temperature and pressure. Had some other novel products for gas wells also. Couldn’t find a buyer in time, the company folded and surrendered it’s IP to the investors.
I worked at Teleco Oilfield Services from 1983-1984 in the R&D department. We were able to obtain 1-2 bps using MUD pulse telemetry in the field. Beside the inclination and direction of the drill bit, we also built versions that communicated resistivity, gamma radiation, weight and torque on bit (up to 50K ft/lbs). A mil spec 6800 running a real time OS (in assembly language) collected data and compressed it before transmission.
Multi User Dungeon?