<img alt="" src="https://secure.leadforensics.com/211043.png" style="display:none;">
August 24, 2021

Projects in Need for More R&D

Season 3
Projects in Need for More R&D

In this episode, Ken and David discuss projects and ideas that should require more R&D.

Featured Episode

Episode Transcription

Hi, I'm Ken Miller.

And I'm David Erdos.

Welcome to the Erdos Miller New Technology Podcast where we spend our [inaudible 00:00:11] time talking about everything drilling tech. Today we've got a fun episode. So we're going to talk a little bit about really big things that should be built. And by that I don't mean a mega structure. I'm talking about a big idea.

Big technologies.

Big technology. Take a lot of R&D money, a lot of time, a lot of trial and error. A lot of whoops, sorry, I messed that up. Let me fix it. Not a project that you're going to be able to just deliver and bring to the market in 18 months. It's probably going to be more like three to five year hauls to like really get it working.

True R&D projects.

True R&D projects. So Dave, you'll take us down the list and we'll talk about them and what it would take and what the benefit would be and how it would change the industry. We'll just have a fun discussion.

Yeah. Sounds good. So first one is downhole GPS. The GPS, as we know, is this constellation of these geosynchronous satellites in orbit. So they transmit at 1 to 1.5 gigahertz, which cannot propagate very far in dirt.

No, it'd be really nice if it did. We'd be kind of out of a job, not quite. We'd still be making pulsars and transmitters, just not directional systems.

Yeah, not direct instruments.

So I tend to think of that as the earth is not transparent to the radio signals at 1.5 gigahertz. If you think about it being either like a wall or glass, it's much more like a wall. Does not propagate. I was thinking about how annoying it would be to live without GPS through the day.

I remember printing out paper directions and watching for street names and whatnot.

MapQuest and everything else like that. Yeah. It's come a little too far, one might say.

Yeah. We've gotten lazy.

But the way that ships used to navigate before they had GPS is exactly what we do for oil and gas right now. It's dead reckoning navigation.

Well, and star trekkers.

Yeah. Star trekkers. So I guess that was a really primitive form of GPS. Yeah. Which is even better than what we're doing right now. We don't even have stars.

We have no absolute references.

Can't look at anything. But they used that for like, where am I in the world? Not where am I exactly, right?

Yeah.

Whenever I'm explain to somebody how we navigate downhole, well, it's very similar to a ship navigating on the ocean. So we start off when we go west to 20 knots for, I don't know, three hours and then go north at 19 knots for four hours, and you calculate your position that way.

So when you calculate that position, you're going to have some error. the error might come from the fact that, oh, well, there was a headwind when we were going west. So you thought you were going 20 knots, but you were actually only going 18. You couldn't quite measure that. And Hey, ships don't turn on a dime, so it took a quarter mile-

Radius, yeah.

... for you to make a radius to turn. So you didn't take that into account. So the problem with dead reckoning navigation is the further you go, the more you're off. There's an error that grows-

To cumulative.

... as you go. The nice thing about GPS, at least I call it, it has consistent referential error. There's some error in the GPS measurement, but it's always the same amount. It might be, well I don't know, a meter or less than a millimeter, depending upon your accuracy, but it's perfect. The more you walk around under it, it doesn't grow. It doesn't accumulate. With our drilling, we're accumulating an error every single joint we drill further. The further we drill, the less sure we are of exactly where we are. So the idea would be to how could we have a downhole GPS? How could we build something that would give us that equivalent referential ability?

Where you would have base stations on surface triangulate to the downhole drill bit.

Yeah. So there's some crazy ideas here. I mean, I'm jealous because the guys that work in the municipal HDD drilling market, they've already got this. Because they're able to basically mag range with AC or DC coils or whatever all the time, because they're shallow enough, 100 feet, 200 feet underground, you can do that. I think we once did the calculations to see how big a magnet would be that would be able to be sensed 20,000 feet away. And the number's pretty astounding. It's a rig sized amount of current and metal. You're frying all the electronics around you and good luck. Not going to happen.

Don't want to be in here with a pacemaker. Those magnetic fields would mess you up.

Yes they would. Yes they would. So, I mean, there are some potential ways of doing this. I think it could potentially be done with a low-frequency EM system. You could have either ME transmitter stationed around the rig. See the problems with EMs, it's so fast. I don't think you're going to be able to put four beacons around the site and be able to triangulate to that. You might need four beacons in Texas and somewhere around the state that are really big beacons. Maybe one in Dallas and Houston and one ...

One Midland.

One in Midland and all the tools are triangling those. But that could potentially work. But that would be really tough because the EM signals propagates so fast. But there is precedent for being able to transmit waveforms less than 100 hertz, basically all the way through the earth. That goes back to the original inspiration for EM systems, which was the US and the Russian submarine communications systems.

Well, the problem with those really low frequencies is your accuracy is also not as good because the wavelength is so long.

You'd like to average it a bunch or it might not work. But that's a potential idea. The other way you could do it as acoustic. So you could consider ... be a lot of math, but you can put three or four big thumpers up on surface like you do for ...

Seismic.

... seismic, and you have to deal with the fact that all the layers are there and all those reflections and be a lot of math, a lot of fun times, but that could work, too.

You could pick it up with an extremely sensitive accelerometer and look at the arrival times of the different signals. You have to have a seismic model to begin with so you can model the propagation through. It'd be tricky.

Yes. The question is, would it only be as accurate as your seismic model? Or could you overcome that? That's the question I mean, I have to think that you could overcome it because we think that, oh, well the GPS guys have it easy because their signal's just going through air. But their signal actually goes through probably just as many mediums as we would doing that with acoustics underground. Because they start transmitting in space with no atmosphere. Then you have the ionosphere and stratosphere, whatever else, all those different layers of the atmosphere, all pretty significantly different in their makeup and their density. So those cause a ton of reflections and issues.

And clouds.

I don't know how they model for that or whatever. But GPS has certainly overcome that.

Although, I think the atmosphere might be more uniform throughout the world than the rock formation.

Yeah. Probably. So I'm just saying it's not vanilla. It's not super easy.

No, no. GPS is not trivial. I mean, I think it's one of the only technologies actively used every day that required Einstein's theory of relativity.

Oh, really?

Because the timing between the satellites is relativistic. Yeah. Relativity comes into play.

That's crazy.

Just the fact that time travels differently at different points in space.

That's a little bit insane to think about. You blow my mind. I did not know that that was required. So how would you build this? You probably get a team, 10 guys, something like that. Several years, you got to build some big base stations.

Some big budgets.

You got to build some downhole tools and electronics, some software, and you got to do a lot of testing.

Probably lots of simulations and modeling.

You need some math experts. I don't know, price tag, what would you guess? Let's make it-

Tens of millions.

Tens of millions. let's play with a scale of I want to say 5, 10 or 25, right?

Yeah.

I don't know. Let's do 2, 10 or 25. Between 10 and 25, somewhere in there?

I would definitely put in that range.

For less than 25, probably over 25, maybe over 25.

Estimating a project like that is ... it could also be total bunker. It could just be impossible physically.

Well, you do what you do. You estimate it the best you can and then multiply by four. Then that'll be an accurate estimate. So it's a big one.

Well, and your feasibility study would just ... are the physics even possible? I'm no physicist. I couldn't tell you.

Yeah, some seismic guy might be like, "These guys are jokers. They don't know what they're talking about."

This is totally impossible.

Total waste of time.

Your accuracy is going to be 300 feet.

I like doing impossible things. So it's pretty good. Well you do 200 feet accuracy with one shot, but you average that over time, that might still be a lot better. But you pull this off and mags go away, gyros go away. We know exactly where the wellbore is. You can space it however you want. You can get as close as you want, far away as you want. You'll know exactly where you are. Code of uncertainty is a thing of the past. We don't have to talk about relativistic probabilities of where the wellbore is. You would just know. It would be far more accurate.

You can drill farther because your error doesn't accumulate with distance.

You can be at the end of a 20,000 foot lateral and you know dead on where you're at. so that would be really cool.

Yeah. I mean, it'd be a massive game changer.

Yeah. So what else on that idea? Or should we go to the next one? We talked a lot about that one.

Yeah. let's go to the next one. Depth.

So depth tracking. Depth tracking right now. This assumes we're going to keep using mags and gyros to navigate. Depth tracking is still a travesty. It really is. What do we do to overcome that?

There's a lot of different errors. We assume pipe is 30 foot or whatever, nominal length. But obviously they're not necessarily the same length. There's going to be variations. It might be a couple inches, might be less than an inch, but those areas are going to build up. You have thermal expansion, then you have elongation based off of the stretch.

You're talking about all the small physics errors too. You're not even talking about the oops, I measured this point wrong, or I screwed up the top drive, the encoder or whatever else.

Yeah. Or we mis-tallied.

Mis-tallied.

There's extra pipe in there that we didn't account for.

With the GPS system, you get depth. That'd be cool. But short of that, I start to think of, how can we get it done downhole properly. I've got to think that it's got to be like a sensor fusion problem. You'd have to make a module that goes down hole and probably uses excels mags and gyros to get some sense of inertial movement. Pressure. Pressure would definitely help you out when you're vertical. When you go lateral, it wouldn't help you anymore. But why not input it? Because ultimates are just so easy. They just work off of air pressure.

Phone has an altimeter that's probably accurate to a couple inches.

It's darn horizontal drilling. If all we had to do was vertical, we'd have perfect depth systems.

Easy.

And then we got to have, I think you got to have some way of measuring traversal of the rock. I think you got to have either some ultra rugged little calipers or you've got to have a-

A little encoder wheel that goes along the formation.

A little ultrasonic something, some sort of laser, like an optical mouse sort of thing that can actually track some physical distance moves. Actually looking at the rock and seeing that.

Reminds me of the Tomahawk cruise missile.

What does it do?

Are you familiar with how that works?

No.

So they preload the topography of the region that it's going to be used in. So it uses its radar. This is before they had GPS, it would use its radar to identify topographic features in the mountains and the terrain to know where it's at. So similar to that, you could track based off of the exact rock surface you're seeing. Same with optical mouse.

But you could also program into the system, you could program in the expected formations and then have it input the gamma sensor readings and see if it can make some sense of that. But a big common filter, big sensor fusion kind of problem.

Or if you could do time of flight up the pipe with some kind of signal to determine where you are.

Yeah. That's the other way we thought about doing it. Full disclosure, we filed a patent on that one. So yeah, there are some other ways of doing it that we're thinking about. I mean the mud pulse could give us a way to measure depth. I mean, accurate timing of mud pulses, if I'm trying to measure the length of a pipe and I go from a starting point to the end, what more do I want that a fluid-filled cavity?

I guess would be better is just a mirror at the other side. So it can point a laser and have it come back. But short of that propagation of a wave through the pipe should be relatively consistent. So there's some interesting ideas about how we could get that done. That's less of a sensor fusion approach and more of a just direct measurements timing approach. Cool. Yeah. So what do you think? How long to build that and how much money?

I would say you could probably do that in a year, depending on which approach. Sensor fusion versus-

That one's shorter than the GPS program.

Definitely shorter than GPS and cheaper.

Two years.

Yeah. One to two years.

You probably need about 5 guys for that one, it doesn't sound like you need quite a 10 person team for that one.

Yeah. Probably not as many physics physical challenges, probably more implementation and fine tune.

I'm still going to say between 5 or $10 million. I don't think it'd be a cheap project.

No, probably not. I mean, if you want it accurate to where you can compete with proper calipered pipes and really good record keeping and tallying and thermal stress and all that.

I want a depth measurement that just, I forget about it. It's just that accurate. It just works. It goes on around. I move the pipe forward. It shows that we moved it. We forget about all this crown and coder depth tracking pipe tally garbage. Just forget about it. So anyways.

Yeah. I think I would agree. A year or two, maybe in the five-ish million dollars.

So that one's one that can be approached and estimated and approach a lot more practically than the GPS one. That's more of a scifi, I wouldn't say ivory tower, because we like to build things here and actually use them as opposed to just building them and look at them. But definitely less academic.

Yes. Less heavy on the research, more on the development side.

Yeah, totally. Going to be half, half, right?

Yeah.

Okay. Good. What's the next one?

True inertial gyro Systems. So modern aircraft, of course they have GPS, but even before GPS, we had these inertial navigation systems that would, and ships have it too, where they would integrate the accelerometers to calculate position and use the gyros and magnetometers and fuse all that data with a common filter or something. You could estimate actual position traveled based off of these sensor measurements.

This is just another form of dead reckoning navigation.

It's still dead reckoning, but you're integrating all these sensors and getting your position based off of that. Now the error with that is also cumulative. So there's some downsides to that, but by fusing all this data, you can probably get better than just pure, dead reckoning based off of pipe depth.

That would be a lot more accurate, right?

Yeah.

I think that it'd be nice to get away from mags. Have a lot less interference, pack wellbore's a lot tighter. Another reason this one's coming up for me is because I think the gyro technology has consistently evolved. There was someone from one of the majors the other day that told me a joke that was like, "Yeah, we kept drilling with our magnetometer systems. And then the gyro guys kept saying, every year they go, 'well, we're still tuning the common filter. We'll see you next year.'" That joke ring very true with me. I was like, "Oh God, that's cringy how close to home that hits."

It hurts.

It hurts. It hurts. But the systems have significantly evolved, right? The amount of modern math and processing and everything else that's gone to these sensors from just the existing manufacturers or what you can cobble together and build today has become pretty impressive. So I don't know about high temperature, but I definitely think that gyros in the next 5 or 10 years could completely take over the Permian.

Under 100 degrees Celsius, I think you could build some crazy accurate record ... Mr. Robert [Estees 00:17:09] has my favorite, I guess he coined ... or maybe he was the first person I ever heard it, said my favorite accuracy quote of all time, which was "dead nuts accurate." I don't know what that means, but it's pretty hilarious.

Very accurate.

Very accurate. Apparently it means very accurate.

Beyond that.

I'm not really sure why. But I think that we can definitely look to gyros for the entirety of the wellbore navigation.

Well, I think the case of wireline jar after the fact after it's drilled, tripping out wireline gyro, inertial navigation, that's probably pretty much solved problem by the gyro guys at this point.

I'm talking about while drilling, moving ahead, surveying, continuous [inaudible 00:17:51] off the gyro, super accurate while drilling. No mags, in steel collars. Forget about non mag. Who cares?

That's the hard part, vibration. The sensor saturation kills you.

What we should have done was our homework and qualified what the value of each one of these ... if we removed all the non mag, how much would that saved the industry? And does that justify the R&D cost or whatever? But that'd be a fun one. Cool. So what do you think? Building a gyro, timeline? Rough order of magnitude?

Yeah. I would say that's definitely in the five-ish.

Really? I was going to go lower.

One to five, one to five.

Well, one of the five is a big range.

That's true.

I bet you my grade will be somewhere between 1 and 99. Personally I'm feeling like it's a two to three year project, so it's a big project, but it's not a crazy long haul.

Well, the question is temperature. Temperature is a big deal. So Permian, I think that brings it down closer to the lower singles range. If you're doing [crosstalk 00:18:54]-

Got to get to high temp.

That's a different ball game.

Why is that?

Well, I mean, it's just the gyro technology in part. If you're going to those lower temps, you can have MEMS gyros, potentially, maybe opt to electronic gyros, like FOG, fiber optic gyros.

MEMS gyros still not as good as a FOG. The problem with the FOG is it's a light driven sensor. So a fiber optic gyro, the way that I understand it works is we have a huge spoil of fiber cable. And we send a light beam down it. Then we look at how much that's moved in the channel as it exits. Is that the correct Understanding?

Yeah. As that coil rotates, it changes the phase of the light.

Oh, it's the phase.

Yeah.

Okay. So it stays centered in the photo conductor, right?

Yeah.

But it's changing the phase. Could take slightly longer or slightly less long to travel.

Yes. If it's traveling with the rotation, it's going to take less time than if it's traveling against the rotation.

So probably what's on either end of that coil, I can probably imagine making the photo coil relatively high temp. But I think the Achilles tendon here is going to be the diode and the photodiode that's creating that light and picking that light up.

Photo diodes are ... it's a very well studied phenomenon, the photo diode noise increases with temperature, which leads us. Noise just shoots through the roof at temperature on the photodiodes.

Totally. [inaudible 00:20:15] photodiodes dead. Could we build it with a PMT and crystal kind of thing, or a PMT? Could we use a foot of multiplier tool?

I don't Know.

At the end of the fiber.

Yeah. So you have to make sure the refraction or the grading. That's complicated. I don't know how the FOGs work, but yeah. That's a good question.

So getting a hyper accurate gyro based system done for 125 C or less, pretty practical, could be approached today. 175 C or greater, question mark on how much money. I'd probably say 2 to 3 million to get her going and then another 2 to 3 million to get it up to 175 C over three to five years, is probably about the right thing, right?

Yeah. Well, let's not forget the filtering aspect because saturation, the accelerometers, is really going to kill your initial solution. So if you can keep the accelerometers from saturating.

[crosstalk 00:21:08] lots of accelerometers. Last thing I say before the last one, but it's hard to predict how long something's going to take in R&D. Unfortunately, very rarely do you plan a three-year project and you go, "Oh, well I got it done six months. That was great. That was easy." That doesn't usually happen. It's usually how much longer than you think it's going to take.

It's never happened to us.

So far the worst one I heard was there was some guys trying to develop some 200 degrees C technology. I was talking to this guy the other day. He goes, "Well, they told me they were working on it when I started here 30 years ago. And we just came out with it last year."

I was like, "Oh my gosh." So it was a 30 year development timeline for this 200 degree C technology. I was like, "Ow."

That's either some really intense perseverance and dedication ...

Or really slow.

Really slow, or just sometimes you got to cut your losses and move on.

We work on this thing every January.

Work on it for a week.

Put a little time into it. So that could have been going on. Maybe it wasn't that focused. But it certainly took them a long time. They talked about it for 30 years. Take us to our last one,.

Last one. So this kind of ties into the photodiode problem of active cooling. So if we have reliable, downhole active cooling, we can achieve other things like gyros that we can't currently do.

So I think this conversation is coming up a lot more now because a lot of people are getting excited about taking existing directional drilling technology and transplanting it into the geothermal industry. I think a lot of existing oil and gas companies are hoping that that market's going to expand significantly and be another market that we can deploy similar, if not the same tools and personnel and practices to. And especially with this idea of closed loop geothermal. Geothermal is all about the gradient.

So instead of having to drill into a volcano and melt your bit to get the gradient, and you can only do this in Hawaii or Japan next to an active volcano. I think the idea is that, okay, well, we can just drill deep enough now or long enough laterals, or whatever, to establish that gradient. And we can do basically closed loop geothermal anywhere in the world. There's a whole ton of companies that are out there competing and trying to figure out that space.

But there's going to need to be a service company ecosystem that evolves to service that market. Basically the bigger the gradient, the hotter, the better. 200 degrees C is a starting point where they can get going. That's still on the edge, even for existing oil and gas tools. I'm going to rant here for a second. But it drives me nuts when we talk about a new technology and someone says, "Oh, well that already existed. Somebody did that in '84. Somebody did that in '91."

It's like, "Okay, great. You know what? Somebody did do that in '84. But they built two and they ran them a little bit and then they, 'Eh, works okay,' or not. And then the shelved it." Whenever I say, "Okay, let's build something or [inaudible 00:24:11]." I'm talking about a mass market available commercial device that anyone can pick up and use, and not some science experiment that somebody halfway finished at Baker Hughes or [inaudible 00:24:23] or whatever, back in the '80s or '90s or whatever. That's what building technology means to me. Okay, end of rant. But these guys really want to get to 300 degrees Celsius.

Just quickly on that topic, it's amazed me having worked in oil and gas for over a decade now, I guess. But just how much effort it takes to go from proof of concept to fully commercial system that is used by hundreds of people.

It's fun challenge. It's a lot of work.

It's a lot of work.

It's a lot more work than you think it would be, right?

Yeah.

I like software better now. Software is easier.

Compiling is cheaper.

It's way cheaper than spending other board rev. So the big question is if we start moving towards 300 degrees C systems. So with geothermal, we don't need any fancy [LABD 00:25:14] measurements. Maybe we need gamma, maybe. Probably don't even need gamma because we just really need to go deep. At least for closed loop geothermal. So I need a temp sensor, I need a pressure sensor, I need accelerometers and mags.

Then we know we're going to have to do some mag ranging to make these closed loops and be really, really tight and consistent. So that's a pretty simple tech stack. That's the tech stack that the majority of the industry works on today. Question is, can we get it to work at 300 degrees C ambient or do we have to make active cooling?

Active cooling would certainly be attractive because 300 degrees C electronics is still very R&D.

Crazy. I hear you, but I have seen several 300 degrees C components advertised off the shelf from certain manufacturers. It is claimed by a company, I don't know how well these things work, but I literally just Googled 300 degrees C microcontroller the other day and something came up. There was an arm seven microcontroller off the shelf that you could buy. It claims that its operating temperature specification is 300 degrees C. That's pretty crazy.

Yeah, that's mind blowing.

Because usually, 300 degrees C has been Sandia DOE territory for whatever. Custom Silicon, all sorts of crazy stuff.

Well even, they couldn't get there until recently.

I haven't seen anybody actually get there.

No, a real system.

There's a DOE report about building a magnetic sensor that they were trying to get the 250 C. I think they kind of gave up about 220. Couldn't really push it much further than 220. the basic idea is you take your 175 C tech, Because we can't cool down the room temperature. I guess you could. It's crazy amounts more work. The idea is we go into a 300 degree C situation and we use active cooling technology to reduce the operating temperature of the tool from 300 degrees Celsius down to 175 degrees Celsius.

So it allows us to repurpose existing technology. Then elevate the temperature. So I guess active cooling is very focused for the [NABD 00:27:24]. We can't really actively cool a motor or a bit. Those guys are just screwed.

Deal with it.

So they just have to get all the elastomers out of there. And they had to use some really exotic metals and just really nice tolerances and just make it work. For NABD active cooling the electronics. So how can that be done?

I mean, there's couple of different ways. There's Peltier technology. So that one only works if you have a lot of power available.

So we're talking about like a 200 watt turbine would barely make a dent in this thing.

Yeah. So Peltier, it's a thermoelectric device where you apply current to it. And once side gets really hot and the other side gets cold.

They're fun to play with. We used to play around with building Peltier coolers for our Intel CPUs back in the day. In high school.

So that's one technique.

Not at all a practical technique.

Probably not practical. They're horribly inefficient. So unless you have just tons of free power downhole, good luck.

Give me an idea, but I'm going to keep it to myself. All right. Keep going.

I got an idea to actually. Maybe the same one.

Maybe The same one.

Compressor. So an AC refrigeration system, you have a refrigerant fluid, so you compress it on one end and then you evaporate it on the other. So when it evaporates, it cools down. So just like the AC in your car or your house. So that's one approach.

Is that similar to what we think the bigger paper was? There was a DOE paper that was published?

No, I think that was different.

Different.

That one next.

So you could build it like a home refrigerator.

Yeah. So you have to dump the extra heat. You're moving heat from one place to another. So you're taking the heat away from this section and dumping it in this other section. You'd have to have your downhole motor and compressor.

That's just the thing is, if you're at 300 degrees C and you're making something else cooler, that means something else somewhere else on the tool is going to be way hotter than three degrees C. Because you have to be able to sink the excess temperature into the ambient.

Because you're just moving the heat.

Yes.

So that's pretty challenging because you have to have your compressor, your motor, you have to drive all that stuff and your refrigerant and complex mechanical system, electromechanical system. So I think that'd be pretty hard to make reliable.

Refrigerators by themselves are just are not super reliable, not well-known to be super reliable.

That just seems complex, but probably doable. But certainly higher on the R&D budget. Then last one that I could think of, and maybe some of the other people can think of some other ways to do it is a evaporated cooling, basically. You have a reservoir of a fluid or some chemical that you were evaporating over time and that evaporation takes heat away. So just like when you get sweaty, the water evaporates from your skin cools you down, so same technology.

Where is it going to go when it evaporates?

You could potentially dump it into the formation or into the drill pipe, or maybe to another chamber. You have a solid mass. Then it could have evaporate and expanded into this vacuum or this other region in the tool that's just like a tank. So as it heats up and evaporates, you're sucking heat away from the electronics.

This is a highly multidisciplinary problem, right?

For sure.

This requires electronics expertise, mechanical expertise, probably elastomers.

Chemical.

Chemicals. I think the only way that you're going to be able to get around the chemical one is the Peltier with a giant generator. Then it's purely electrical mechanical and electronics. But you need a very diverse team of very talented individuals to pull this off.

You still probably need all of this in a flask.

Really a flask?

I think so. Because depends how long-

You want one flask to thermally insulate whatever you got going on. Why can't we just put everything in a flask right now?

I suppose you could, it just depends on how long you want to drill. So if you're okay drilling 20 hours at a time, tripping it constantly.

That's the downfall of flasks as they are.

It'll eventually heat up to the ambient temperature. All you can do is slow it down. So if you're tripping in wireline taking some measurements and going in and out, you can get away with it.

Well, the basic idea of the flask is that you create a vacuum between the electronics and the outside. Heat just very slowly moves across the vacuum. Way lower than being filled with air or whatever else.

So the vacuum just acts as a good insulator.

So the reason we can't just take a flask right now and flask up an NBD tool like you do for the wireline tool is because NAB runs are on the length of 50 to 100, 150, 200 hours sometimes. The flasks just can't even keep anything cool for a 10th of that time. It's just not practical.

Now you might be able to do, let's do your tool's ready for 200 C and you have a 230 degrees C well, you might be able to do a flask to get you enough time ...

To get past that. But the reason the wireline guys can get away with it is because their runs are shorter. On the orders of several hours or tens of hours in and out scanning the formation. So that's why flasks are just a fantastic technology for wirelines.

Saves them a lot of effort on the temperature.

Makes sense. You thinking 2, 3 million, 5 million, 10 million, 15, 20?

I think more in the five.

100 Million, $100 million dollars.

Yes. Probably in the 5 to 10. Depending on which approach. So I think the evaporator is probably the cheapest, I would say. The compressor base, refrigerant loop based cooling, probably pretty high up there. Then Peltier is also pretty high up there, I would think.

I'm going to say three to five is my guess.

Three to five? For which one? All of them, any of them?

I don't know. One of those to get you going.

That's about right. Yeah.

A couple of years. Two, three years.

Yeah.

Should we be relatively approachable. Should be able to have a measurable start and an end. It's going to be tough, but it's not going to be as much true R&D.Yyou might have to build several prototypes of each one and see which one has the most potential as a feasibility study or exploratory study, and then go from there.

I think materials are going to be a challenge. Seals and chemicals and all that. But it's a very understood problem. I think we understand thermodynamics and heat transfer pretty well these days.

Yep. Now you gave me two ideas for further podcasts. One, I think we could make a podcast about commercializing technology because that'd be pretty, pretty interesting, how you go about that whole process. And then two, all these price tags, we haven't really hit anything that's the $25 million price tag. Everything we've talked about today is probably less than $10 million per project.

The downhole GPS could go to-

The downhole GPS, the big one. Okay.

If it's even physically possible.

Boundering on if you could prove on a computer that you have a chance in hell, it's going to be a big amount of money.

That one I could see exceeding its budget by a good amount.

If only rock was more-

Transparent to radio.

... transparent to radio. It's just such a disappointment.

Make our lives a lot easier.

It would. Even the underground, what is it? The ground penetrating radars only goes 300 feet down maximum.

Not very far.

But now I'm thinking we need to think about projects that are $100 million dollar projects. Big ones.

Astroid drilling.

Asteroid drilling.

We'll leave that for another episode.

That's good.

All right. For any new listeners, please check out our podcasts on iTunes, Spotify, and YouTube and like and subscribe. Thank you.

All right. (music)