Insider's Guide to Energy

44 - Fixing the methane leakage

October 31, 2021 Chris Sass Season 2 Episode 44
Insider's Guide to Energy
44 - Fixing the methane leakage
Show Notes Transcript Chapter Markers

This week Chris and Johan are joined by Adam Kingdon. He founded Utonomy Ltd, which at its heart addresses the methane leakages in the gas distribution networks. With a combination of specialized hardware and IoT they can drive down costs for their customers. The three discuss physical and cyber hazards associated to this critical infrastructure and how machine learning is implemented in the system. Listen in, to find out how the company optimizes biomethane distribution and how it’s adapting to an ever-evolving market. 

Broadcasting from the commodity capital of the world, Zurich, Switzerland, this is ‘Insiders Guide to Energy’. 

This edition to Insiders Guide to Energy is brought to you by Fidectus. Go to www.fidectus.com for more information.

 | Timestamp | Speaker | Transcript

| 08:08.53 | chrissass |  welcome to insider's guide energy I'm your host chris sasse and with me as usual is Johann oberg yohan how's it going this week
| 08:20.10 | Johan | It's ah another week chris great to be on the show. My week is as a normal week which means I'm extremely jealous of you this week to be able to go to my hometown. Well at least my second hometown so I'm more interesting in understanding.
| 08:35.26 | chrissass | Well I'm back in London I've been back here in 3 weeks I think this is a second time here in 3 weeks. So it's nice to be being able to travel and see my customers again and be you know talking to energy traders face to face as opposed to doing it on facetime or doing it on some video conference.
| 08:35.42 | Johan | How's your week.
| 08:53.65 | chrissass | Ah, it's great. Um, it's nice and warm here. Although it was kind of a great a we did see a little bit of sun. It's it's great to be in London it's it's a lot of fun.
| 09:01.57 | Johan | Fantastic place as I said bit jealous but another week another show another guest looking forward to this 1
| 09:09.13 | chrissass | Well, interesting to be in the u k with our guests right? So you know what we were watching is a whole bunch of noise and and information about gas gas prices shortages supply and so although we're not going to directly talk about that. We are going to talk about gas and our guest is an expert in this he he brings some interesting technology and so when he and I first talked in a pre-conversation. Um, he talked a little bit of internet of things kind of stuff and and and pressure reading and gasline and stuff been in it. First it didn't click to me of how this helps with renewables and how this helps helps with the whole story of what's going on and so I'm hoping through today's interview we figure out how having all these sensors and all the pieces that he makes with his company brings value to our customers and our audience.
| 09:52.23 | Johan | Um.
| 10:01.39 | Johan | Now I agree and I think for me this is especially interesting because it kind of connects my previous role with my current role. So so you know as you know out exactly Ah, not that far back. But you know that we are with with our team in Germany we work a lot on biogas. Of course we're in energy.
| 10:11.49 | chrissass | Mcdonald's.
| 10:21.00 | Johan | Company. We're an n trader so we we do some of the gas as well. But in my previous role at tellia we built the iot business. So so the iot has always been fairly close and we we launched products as the iot in a box with with a sense of sensor pressuring kind of the software together with hardware combination. So I'm I'm really curious to learn a little bit more about how they actually put this into scale and and actually create value from what used to be back in the days only a few years ago technology only so this is really looking forward to see the real value coming out of this now from these kind of technology themes.
| 10:57.37 | chrissass | Well what we can do is we can speculate or I can just introduce Adam kington ceo of autonomy limited adam welcome to the program.
| 11:06.64 | Adam Kingdon | Um, well thank you very much for inviting me Chris It's a great pleasure to be here.
| 11:10.39 | chrissass | What we're excited to have you as our guest is you can hear in our pre-conversation. We we anticipate where you may go. But as as Johann is apt to say is we don't really know where the conversation is going to go but we have some expectations right? or wrong. But before we go into the technology. Maybe you could start by telling your audience a little bit about your background and who you are.
| 11:31.59 | Adam Kingdon | Well, my name's Adam king I'm the ceo and the founder of a company called euutonnomy and I founded deutonnomy back in 2015 ah very specifically to develop new technology to reduce the leakage on the gas distribution network I'm sure you're aware that. Methane leakage is ah is is is is um, a real serious worldwide problem and and you know particularly because it's so much worse greenhouse gas than c o 2 and but but proud of founding euutonnomy I found a company called ito water which. Actually developed quite a similar technology. But for the water distribution network to reduce the leakage on the water network and and while I was running I 2 o I had a call from a couple of gas Networks saying yeah, we also have a problem with leakage. And maybe the ito o technology could help us as well. So so we had a look at the the um we look at what you know and gas networks were a lot different from water networks. So we end up deciding to stick with water but when I left I 2 o in end of 2014 I got back in touch with those gas networks and said. Ah, you still looking for a solution to help you reduce your leakage and they said yeah you bet we are. Yeah this is a really important issue for us and if you can come up with some some ah clever technology and better way of doing it. We'd be really interested. So so that that's that's what I founded euut autonomy.
| 12:59.88 | chrissass | So so when you say leakage um, can you help me understand what that means and and what causes it and where it's going. Is it just going to the atmosphere. What what's happening for the leakage.
| 13:13.54 | Adam Kingdon | Yeah, well well, it's so some of the some of the pipes still in the network in the older older cities is is still is cast iron or ah and and each and there's not a huge huge ah percentage of gas lost. But each joint leaks a little bit of methane and and across a whole network of thousands of kilometers of pipe that that can add up to significant amounts and and that that methane is just released into the atmosphere and obviously methane is terrible greenhouse gas and. Yeah, over a twenty year period it's it's ah it's about eighty four times worse than c o 2 um, obviously people usually usually talk about 100 year period where it's it's about 29 times or 28 times worse greenhouse gas and c o 2 but actually I mean given that we've got some. We're really worried about the short term and we're really worried that we're getting to maybe getting some sort of tipping point and this there's quite an argument to use the 20 year figure of eighty four as certainly as much as the the 29 figure for over 100 years
| 14:23.97 | Johan | So It's ref find if I if I'll go back a little bit to to Chris question in terms of of how it actually works and and probably a stupid question. But obviously you you've put on some kind of sensitive data around the the network and and that triggers where the leakage is and. How does how does it actually works and how do you actually repel it or how does this.
| 14:43.17 | Adam Kingdon | Well, yeah I should have explained that. Yeah we we? What? what? we're actually doing is is optimizing the pressure in the gas distribution network because the the leakage is directly proportional to the pressure. So if you if you could reduce the pressure on average. Over the year by say twenty percent you're going have twenty percent less leakage basically goingnna lose less gas through the through through the existing joints and leaks and that that exort exists in the pipe. So so we don't actually fix the leaks but we reduce the amount of gas that's lost out of the network and. And and managing that pressure is actually surprisingly difficult job and and and that's why I mean that's why the haven't been solutions are solution that and hasn't been used before It's a very tough thing to do so.
| 15:31.55 | chrissass | So you you talked in the pre preinter interview with me about about the managing the pressure and I think you you mentioned that they send cruise out to actually adjust and and in in certain networks because it was hard would only do it several times a year they weren't actively. Monitoring the pressure. They weren't actively changing is that do I remember that correctly from our conversation.
| 15:52.15 | Adam Kingdon | You yeah well that's correct because the I mean the pressure that you you have in the pipes say in the city or a town is controlled by thousands of gas governors. So so the the the gas flows from the higher pressure tiers. Through a gas governor into the low pressure tiers which are low pressure pipes which distribute the gas to the houses and all those gas governors are fixed have a fixed outload pressure I mean they're completely Manual. So if you want to if you want to adjust the pressure output you you go along there with us. You hope. They're sitting in a kiosk side of the road you open open the door you go and there with a screwdriver and you adjust the pressure by hand and and the problem with that is that the obviously demand varies hugely throughout the year it's on a cold winter morning. You may have a hundred times a demand that you get on a warm summer evening and if you're gonna gonna have enough pressure. Just. Deal with the cold winter morning then you, you've got to set the the the Governor's pressures very conservatively high and that means that through the Summer when you got less less demand was at night or the summer you got much lower demand. Yeah, you've got to you got. More pressure than you need you and know lot excess pressure is creating excess leakage. So ah, yeah, and and and the bit you mentioned about manual control. Well they the way they get over the get over at the moment is they have seasonal settings so you know they have ah the Governor's being on a winter setting. And so they be set for pretty high to make sure they can cover any any extreme demand conditions in the peak winter Peaks whenever the first thing in the morning. Everyone's central heating goes on. There's a big peak. they' got avenue pressure in going into the network to cover that. Um. So yes, and then when it comes to springtime things warm up a bit. They don't need that pressure so they send the whole crew around thousands thousands of governors turning the pressure turning the pressure back down again you know and and then if if they're not careful. They turn the pressures down and then we have a cold snap in in April something. Have to run around quickly and tell them all back up again and and it can take 1 network in up to 3 weeks using all their technicians to get around and do this job. So so so the the first thing autonomy system does is is is enable them to remotely. Control all the governors in the network. They they don't have to send people out to to go and do it.
| 18:23.00 | Johan | so so so I think that was that's very I think that's really interesting because the first 1 is obviously then to turn it off and on which which could be done manually or through a sensor or or automatic thing. But then the if i.
| 18:29.60 | Adam Kingdon | Um, yeah.
| 18:39.72 | Johan | Read it correctly if I understand it correctly obviously an an ai or an internet of things does more than just on and off. So so what? What are the other layers that you add to this apart from weather or kind of seasonal.
| 18:44.13 | Adam Kingdon | Um, yeah.
| 18:51.52 | Adam Kingdon | Yeah I think you're very right right to look at it like that because that's how we look at it and how our customers look at it. We. We got the the basic system enables you to remotely control the the governors and then on top of that we have ah we have ah so Cloud host is software which uses ai. Then to work out. What is the optimum setting that each each of these governors should be set at and and it's it's it's a it's a really complicated problem because so you have a town like southampton where we're based It's got 2025 governors all feeding the same network. And they all they all react against each other and the demand is changing very fast and it's changing a lot and and then you you're trying to monitor a number of different points in the network to make sure you don't go below the minimum. So if you go below the minimum then you you risk someone's boiler stopping or cutting someone off. Which absolutely absolutely got to avoid at all costs. So. So so so it's a really tough problem and we we. We've got some really clever ai people on the team and you know we've come up with a we think's very effective, good solution to that. But the hardware was a. Really serious challenge too. Um, and we've had to in enable to in order to implement the solution. We've had to develop the hardware and the software. So the hardwares is really really tough. It's not just an off the shut some off-theshellf actuators and and and route routers. it's routers it's um it's um we've had to develop the hardware ourselves because to be able to control the Governor we have to retrofit some hardware onto the the governors themselves which are seeing in a kiosk and there's potentially leaking gas so potentially leaking methane so you have to regard it as being an explosive environment. So so we have to. Have our hardware certified to the highest level of what's called atex. Yeah, which means that under no. So no circumstances at all. No no, no failures no shorts, no fault of any kind could cause a temperature rise in any of the components which would then trigger an explosion. So. It's a really tough thing to do. And but before we started we found some papers that said, you'll never be able to get ax 7 with electric motors. But we we got some smart engineers and we've got some very found some very very special motors and we've be able to do it and and then you also got no power there and you got no coms and. Anything goes wrong at all. It's got to fail in a safe pressure. Ah you can't have it fail. You can't have anything go wrong and and and and and either well certainly can't fail with ah with an over pressureure because that that that would be very very that would really risk safety.
| 21:44.65 | Adam Kingdon | And you don't want us to fail and cut. Anyone's gas off either I mean that's that's not a great thing to do either. So so so we have to have a lot of redundancy and a lot of failovers. So that if any any any components fail software hangs anything like but like that it it fails in ah in a safe known pressure. Ah so. So there's a lot There's been a lot of really tough engineering going into the hardware as as well as the software to develop the whole solution and ah.
| 22:10.64 | chrissass | So you you talked about the hardware you you? There's a little bit. You haven't yet covered the software angle of of what you've had develop um and so are gas companies going in and retrofitting then for for. Reasons of just loss are they is it because they're worried about Greenhouse Gas is it because of cost of losing gas or what what's motivating them to put your system in place.
| 22:42.51 | Adam Kingdon | Well each market is a bit different but ah in in the uk um, the regulator off chem gets some some pretty pretty um, pretty generous. Well I hope you wanted to accept accept them as being generous. But. But but you know reasonably high incentive. There were incentives to to get the leakage down and and and then in the current price review period yeah they've but they've also they've mirrored that by by. Um, they'll impose fines if if it goes the wrong way as well. So so they're quite. Networks quite incentivized financially they also they also have to publish their leakage figures and they they've they in the price part of the price review period was to set a target for each year during that review period and they and they every year they have to publish where they are against their target. And and and they really don't want to be be seen to be missing their target. Um, it's very very bad reputationally and so and then and then the I mean obviously the the cost of the the gas price going up like it has I mean any gas that you lose through any form of. Leakage or is going to cost someone 5 times as much as it used to cost them. So so that's that's another incentive to to get the leakage down and.
| 24:06.39 | chrissass | So now is the gas coming from a central source or how how is the I understand power grids pretty Well I'm I'm not sure sure I understand the gas so I get that there's a pressure So I'd equate that to like a base load that you'd have to have there that you want to have it where does the gas come into the system and is it. Your system involved in that.
| 24:27.30 | Adam Kingdon | Yeah, we yeah, we're not, we're not yet involved in in higher pressure tiers I mean I I mean basically the the gas I mean suddenly I know the u k better than our other networks but that they're pretty pretty similar in concept. So the gas will come and come in from the. The North sea. Although I think about only about half our gas comes in North sea now and then it's transmitted down the country in in big big steel pipes at about 70 bar which are the transmission lines like like the 4 hundred kv or whatever transmission lines electricity. And probably a bit like electricity that it then goes into ah lot. Love so low tip pressure tier I don't know probably about 20 bar 7 bar then down seven bar and to 2 bar and then and then and then in the in the in the the pipes in the cities is milli Bars. So so but all that. Ah, the the other reason they're going to need control is because that's like like the electricity grids that is starting to change that used used to be the gas just came in at 1 end of the network and and then it fed down through through the different pressure tiers and came came out of came to be out of. Into people's houses through their meter and but but right right now the gas networks the the gas supply is getting more distributed and particularly the first thing we're seeing is is ah a lot of biomethane plants. There are over 100 biomethane plants have been built recently in the uk and they tend to be built in the middle of nowhere. You know they tend to be built where you've got the feedsock so it's in ah in the middle of North agricultural land piece of um in the middle of the farmland and and they have to connect into the nearest bit of grid the nearest bit of grid. Maybe the Mediumn pressure. Maybe it's so twobar or in the uk it's usually the twobar. Network they feed into and so so so so it's all working in a different different direction and they need control to be able to manage that so that's that that's the second big application for the autonomy system is to provide control in these median pressure networks so they can allow allow the feed in for the biomethane plants. And these biometane plants have you also have a particular particular issue with the feed-in because ah, ah if if if something the demand drops on the bit of parlor network where they're feeding in then they can't get the gas in anymore and they've got no storage on site. They end up having to flareir which is. yeah hugely wasteful so so they so the yeah yeah the the networks are we're looking at putting our system into this media pressure networks so then get control of those median pressure networks remotely and and also we're looking at ai solutions on top of that so they can automatically.
| 27:15.18 | Adam Kingdon | Ah, prioritize the biomethane feed-in at all times so that so any demand that is on that bi the network is fed first by the biomethane plant and and the the natural gas is there as a backup and is only used when when the biomethane plant can't can't supply enough. Ah, the Demand. So.
| 27:34.40 | Johan | So 1 of the things when when I worked with Iot in connecting things in 1 if it was connecting the car. So if it was connecting the homes or whatever it was connecting you actually took from usually an innovative. Smaller company, a startup company that had a very good idea that had a solution to a problem but there was always this crucial thing where connecting another and third party software which might not be ah, pressure Tested. It has a direct connection to something that is business critical. In this case, the gaff into our homes even did was this how does the the gas companies look at this where where they're actually allowing if I understand it correct that your software to be connected directly on to. Distribution or or almost managing this.
| 28:30.60 | Adam Kingdon | Yeah, well I guess there are a couple of things I mean firstly the way we've designed our system means that um, there's also local control of the Governor. So. So if if if if the equipment of the Governor loses connection with. With the with the central servers or the the the cloud-based software. It's not as there's there's no security concern at all. It carries it carries on it knows what to do and and and then when it it just doesn't upload its data and it doesn't. It doesn't get updated with with the with the new with new schedules. But it it. It can operate in a safe mode until it reconnects again which means that if the means that if the if the coms fails of the mobile phone Network goes out. No safety issue at all and and and likewise if there's um, yeah, it can it can function without the central software. So if if for example, there's an outage on the cloud server. It would everything it would continue to function safely so that is really part part of the design. Of the system. But but but you're completely right? The um yeah, we're doing a pretty key important pretty key job for the networks and and they they'rere taking a very strong interest in in the software and the way it's been built and and particularly the cyber security is ah.
| 30:02.73 | Johan | 8
| 30:04.79 | Adam Kingdon | Is a massive issue and and they you know they they they do a lot of due diligence on on this on the software and although were where we we've um we small we we got a relative we on engineering Team. We around you know, close to 20 But even so we we we haven't built every bit of the software platform ourselves. So we've partnered with a big software company software Ag which have ah already have an iot platform and we've been able to build our applications on top of that. So. So so really, we've been able to use some. So something like a middleware to does all the basic functionality of managing the devices and and managing the databases and etc etc and and managing logins for customers and we've built our application on top of that. So So we've got a very very robust platform. Um. This's very well proven and sported by very large company that that we've made that we've running our software on top of and I think that helps a lot that's given some confidence to to our to our customers and I we.
| 31:05.80 | Johan | See.
| 31:14.36 | Johan | which which I think and I think this is this is an interesting topic and we discussed this back and forth there' 1 thing of connecting we had a few shows ago a while ago on the show talking about the colonial pipeline obviously and then I right out how that was connected and and we've got some of the other things I'm always interesting.
| 31:14.59 | chrissass | Um, John.
| 31:33.90 | Johan | 1 I have no problems with the actual technology because I think that's usually very good. The quality of the service I'm I'm always a little bit interested in terms of how the larger companies who are operating this business critical part is addressing a smaller startup. Because we discussed that on the show a little bit in terms of the the relationship between let's call it the old legacy versus the the yeah the the smarter or the the new startups that scales up the the new technology and how that interacts sometimestime clashes and sometimes interacts. So.
| 31:56.90 | Adam Kingdon | Yes, yes.
| 32:10.29 | Johan | That's a little bit also quite interesting for me to understand how this works.
| 32:11.40 | Adam Kingdon | Yes, yes, well well of course the ah, the other thing to say is that that we've worked in very very close partnership with our first customer sgn. Ah, you know we'd never have done it without that really close partnership I mean almost from day 1 so so so there
| 32:24.71 | Johan | 2
| 32:30.89 | Adam Kingdon | You know they, they've put a huge amount of amount of effort into this as well and and and we've worked very closely with them to to make sure make sure I mean with ah with a lot a lot of different points of view and and some some very some real experts on on Networks and how they operate and. Ah, governors operate and and safety or the safety issues. Ah, we've we've worked very closely with them to to ensure that that our product yeah is is is is low risk because it is it possible to be for for their network and and they've also got. Um, the customer is sgn is our first first customer on the the networks in the uk they've also got a very strong it department very strong it team and they know this is a way the world's going and they have a cloud first strategy for example, so they you know they've already worked out what it means to put put software safely in the cloud. Ah, without without without risk. So so we're working very closely with our teams as well as has helped helped a lot but but but I mean right, right? right? from the day 1 we've been very aware Cyberurities is Goingnna be a massive issue and so so that that.
| 33:24.33 | Johan | Yeah.
| 33:41.41 | Adam Kingdon | And every part of the design has been. Yeah, we've been take take a note of that.
| 33:42.59 | Johan | Yeah, so so so so early on would you would you say that your your kind of prime customers almost need to have this Cloud strategy in order to make this work or.
| 33:54.43 | Adam Kingdon | Well, it's It's just a really efficient way of doing it. Um, but but fortunately we've partnered with them with a company that's with so software Ig We we can they have got the capability of taking the software taking the the Cloud hosted software. And either putting it into a private Cloud or on on On-premise for the customer. So. So so I mean if they if they if the customer really really wants to have their software where they can see it and touch it on their On-premise then you know we can do that. But it's it's.. It's probably less efficient model and.
| 34:32.15 | chrissass | Sure. So So you you talked a bit about ai and the use of ai you talked about this the software ecosystem and this partner you work with so ai is a little bit of a buzzword these days it means a lot of different things to a lot of different people. So Maybe you could walk us through what. What kind of ai or machine learning or what? what?? What is the system actually using ai for what what is it getting smart about and what is it doing what kind of things.
| 34:58.64 | Adam Kingdon | well well I guess that there there are a couple of um things we need to we need to know to to know what what? how to set the pressures and the first thing we need. We need to understand what what's got to happen to demand over the next twenty four hours. So so part of the ai or machine learning is is is is getting ah as accurate as possible forecast of demand on ah on a pretty much minute by minute basis over the next twenty four hours because that affects that affects what the pressures we got to set ah that affects how we set the pressures. The other thing we've got to know about is is how how how changing the pressure on each governor affects what's happening in the network so you so we also we also develop real time network. We. We also develop models of how the of the network and we combine those 2 things to. By combining those 2 things and we can. We can predict what what is going to happen in the network and therefore what what settings we need for each of the governors. So so it's probably more more machine learning than anything else. But.
| 36:12.28 | chrissass | And then so is it I get that the the weather or the prediction capability Demand has inputs there that the flow dynamics are understanding what the pressure in the the network would do. Is that something that you need to continually learn does it does a system change its behavior or is it like physics. Is there certain laws that are just kind of consistent in in how you expect it to behave.
| 36:37.74 | Adam Kingdon | Yeah, well, Ah yeah I Probably shouldn't go into too too much detail here. But yeah, it's um, it does it does change that Also so also changes over Time. So So we do need. Yeah, we need continuous data from the network to up to update. Update all the models.
| 36:57.48 | chrissass | Cool. So Another thing you mentioned is is you've guys have invented some unique or engineered unique switches or devices to to governors and and to control Things. So What's your supply chain like I mean how do you get economies to scale if you're building these you know are you building them 1 off or do you have them produced in scale as you are you growing or how does that work.
| 37:23.52 | Adam Kingdon | Well, we've we've partnered with a a very strong subcontract manufacturing company. Yeah with with ah you know, ah capable of of producing the product at at ah, ah, quite some scale. So so um. Yeah, know we we don't do any manufacturing ourselves and so so they've been supplying so relatively small quantities so far we've we've just equipped a couple of networks for sgn which we've been running for last couple of years but about we're just on the cusp of moving into into sort of commercial phase of the company. The commercial life of the company. So so we're just working with our with the supply chain to scale up. Um, ah to much higher quantities and so that's that's I mean that's all going, but that's all part of the process of of moving. Yeah.
| 38:08.67 | chrissass | How how how how many of these does a network get are these like every quarter or mile are they ever What? what? what? number of sensors are in the network. These hundreds are these thousands.
| 38:22.67 | Adam Kingdon | Well well there yeah thousands. Yeah I mean in uk if you if we if you covered the whole the uk there'll be about 40000 devices to control governors. Um, there are 20000 government governors and so um. So to control the Governor control all the governors you need about 20000 of these devices and then you also have need sensors out on the network that's measuring that are measuring the pressures in the network as well and sending that sending that pressure data back and the probably similar number of those. So yeah, so so's it I mean for for I this for. it's it's it's not a huge number of endpoints for some iot applications. But it's ah it's it's still quite a lot for us to manage.
| 39:08.29 | Johan | But if it's if it's 20000 total which is kind of the the the overall and I know you're in the startup and this and kind of the the initial stages of this. But obviously you don't do 20000 at once. So what would be kind of an average if it's possible to say in this early stage. But what will be an average project if you go in the city of Southampton. For example, you mentioned a live in the beginning. What will be the what is a normal project if it if it if it there is a normal.
| 39:31.17 | Adam Kingdon | Um, yeah. What? what? we? yeah? Well if the first the the first sort rollout stage. So. There's this Beyond pilot we we're just finalizing the moment with our first customer and that's gonna be around 700 systems so that's the sort of.
| 39:51.99 | Johan | At that system same thing as governors or okay, yeah, not.
| 39:53.63 | Adam Kingdon | So typical is sift of same things. Governors. Yes, yes, you need every every governor you need us. Ah yeah, and and that that that would do um I believe about about 50 or sixty Networks so 50 or sixty towns or or worth of. But the worth of governors. So yeah, so it's a reason it's a reasonable start I it's good starting at ah at ah at a reasonable quite a reasonable scale. Um, and we've been running running pilots since being here 2019 at at at a much smaller scale. Yeah.
| 40:14.46 | Johan | Um, hit.
| 40:25.90 | Johan | So So it's out of curiosity and so out of Curiosity. You know I put my business hat on and then kind of drill in a little bit to to the to the business case in this 1 What where are you making the money is it on the the hardware is it The kind of an operational part is it the software. Or a combination of both I guess in turn initially that it's It's an upfront cost of the hardware. But how does the business case look like for you guys.
| 40:54.10 | Adam Kingdon | Well or for our customers. It's pretty good. It's a we tend to get 3 ah payback? Yeah, um, yeah for yeah for for us we we make us a normal manufacturing margin on the on the hardware that we we we sell the hardware as as a capital sale and then.
| 40:57.75 | Johan | Um, that I am yeah.
| 41:13.67 | Adam Kingdon | We supply the software as a service so then the software is supplied on ah on ah on a sort of annual or monthly fee basis sort of well we we well? Well, we've we It's interesting. Interesting idea you know we've.
| 41:20.50 | Johan | So you don't take a rev share for all all the bunny you stay.
| 41:32.30 | Adam Kingdon | Yeah, there may be clients who who who be interested in that. But and the moment we take a fixed fee for this for the software.
| 41:37.46 | Johan | Yeah, on.
| 41:39.65 | chrissass | And so where does this software evolve to so you you get in a bunch of Data. You're going to have as you go your customer base a pretty big historical database I Assume at certain point. Um, where do you see this going. Where does this change the industry or what does this? do you know where is your company helping drive. Industry.
| 41:59.53 | Adam Kingdon | Well I think in the number number of different directions I mean I I mean first the the first problem we've we've focused on is leakage but then very quickly our customers said well actually we've got this bigger problem which is how we got these biomethane plants feeding in and in the future we're going to have Hydrogen. How do we control the network. To enable that to happen. So so so that's that's the next problem we're addressing um and and and then there's also so so really moving from leakage into hydrogen and bimethane and then also there's ah, there's there's a lot of other problems. We can solve using using the data and ai from the network and there's huge scope to do to do that and so so really, the ultimate ultimate goal is to have a much more automated as far as possible maintenance free Network that's pretty much running itself most of the time and that that's absolutely achievable.
| 42:55.86 | chrissass | So are you are you unique or their competitors I mean you you talked about a fairly long development cycle. You talked about some intellectual property that you developed. So do you have competitors in the space.
| 42:56.66 | Adam Kingdon | So and I mean that's the direction we're heading.
| 43:11.91 | Adam Kingdon | But very little um, that there's there's 1 uk based competitor that developed their their solutions or the end of the 80 s beginning been in the ninety s and there's still quite a lot of units there. A lot of units and in the ground still um, but but it's ah. It was developed way before anyone thought of using machine learning or ai and and and also it's it's quite difficult to install. You got to replumb the governor station and and and it's quite expensive to maintain so so um so with with the first that we know of. Modern technology doing this addressing the solution with the sort of Modern iot technology doing this solution and also using these actuators which are ah very quick fast to install in the Governor stations that you don't have to replub the Governor station. You don't have to switch off the gas you you just. You retrofit retrofit all kits into station and if in a couple hours
| 44:11.18 | chrissass | So is the gas market going to decline as we more electrify the world and and so I heard you say Hydrogen Um is ah is a possible future use but is it a declining Market. So I mean is it going to go rapidly downhill From. From being inside the industry because people don't want to Burn Burn Fossil Fuels or or gas.
| 44:32.98 | Adam Kingdon | Well, yeah, yes, I'm I'm I'm I'm I'm sure there'll be the you know the gas demand will go down as as as as as electricity takes over a larger share of heating in the home etc et Cetera. so so so I'm sure we'll see that. But I think we'll see it. We'll see it at the same time a switch to to low carbon gases so to a big switch to to hydrogen and a switch to Biomethane. So so that'll present significant opportunities for us to because although they only need the Network's going to need. Much more control in the network to enable them to to to use to bring in these gases into the network and so where whether there may be fewer Networks or there may be less gas transported. We think with the the switch to buy metthane um and hydron will will keep us pretty busy.
| 45:11.70 | Johan | Are.
| 45:28.72 | Johan | So that's from a technology or or a business point of view. You're based in the uk you you mentioned southampton you mentioned a few uk cases. Do you also review? geographical expansion. Is there gas markets interesting elsewhere or I know u k is quite big for gas. But.
| 45:51.25 | Adam Kingdon | But but very very much so I mean ah technology is is applicable to any gas network really ah around the world and um and we've we've had some really strong interest particularly from the us particularly for the northeast us where where they have some old low pressure systems. But they still still using it' a metallic pipe and and they also have they also worried about their methane leakage and ah are increasingly coming under pressure to do something about it. So we've had some really strong interest from from the us um, and also from europe I mean even. From countries where they they have more modern networks. They so they still get see value in be able to adjust the governors when they want to remotely. So yeah, no, we we definitely see a very big worldwide opportunity. But but we we we want to do our first rollout at scale in the uk close to home with the customers. We know well and but but very soon after that we expect to move to to the us and to all other other european markets.
| 46:57.39 | chrissass | So are you developing additional sensors or or internet of things that go with the space. So I mean it sounds like you've got governors. It sounds like you've got pressure gauges of some sort or or your r and d or your engineers working on new products or new value ads in the gas. System or is it all based on software going forward where the the higher value comes.
| 47:20.14 | Adam Kingdon | Well, it's It's largely software I mean I mean once once once you've got a hardware installed in the network then then I'm a collect able to control the network we're able to collect data for the network then it's it's largely a quest of developing ah additional further software applications. To do to do new things for them so we may need the odd additional sensor for example to to if we want to measure flow. For example, we might need need to Inert additional sensors but but but it's going mainly software applications using the Data. We're already. We've already got. We're already collecting and and controlling the assets that we're already controlling.
| 48:00.65 | chrissass | So is your team more physical like what kind of engineers. Do you have on of your 20 team members are they software and like engineers are they.
| 48:08.71 | Adam Kingdon | Um, was a whole I mean that's the we got a whole all disciplines I mean that's that's that's the challenge of developing our solution and we had to have some clever people doing a whole range of different things. So we've got we got some really smart mechanical engineers because we had to solve. Really tough mechanical engineering problem we we and and and we're going to meet go to need mechanical engineers going forward because every time we go in the new market. They've got a different type of governor and we have to develop new interface for the Governor. so so mechanical engineering iss on a remainer skill for us. We've got electronics engineers and um particularly electronics engineers who know about atex are quite rare. So we've got a very strong electronics team. We've got a software team that has developed the the software on the devices in the governors and they also develop the software and on the cloud cloud platform. And then also we got a ai team developing the the ai ai solution. So and then we've got a fantastic cto who who you know who's but been responsible for pulling that team together and also leading that team and and coming up with the solutions.
| 49:25.11 | chrissass | Um, did you have it sounds like you've built ah a pretty complex team to to solve a complex problem to did you have trouble finding investors that.
| 49:26.76 | Adam Kingdon | And and and it really has been a challenge So a lot to kill a lot the lot to do.
| 49:37.11 | Adam Kingdon | Um, yeah.
| 49:41.34 | chrissass | Shared the vision to go along and saw this as a problem statement to I mean you're fairly young company. So I don't know if you're a private company or if you had investors come along. Did you have trouble selling the idea not to your customers but to investors.
| 49:51.30 | Adam Kingdon | No, no, no, not at all. Yeah, no, the or the the investors um, always get it pretty quickly. I mean they can see that's ah it's a massive problem and and we got a solution to it that works they know they the investors really get it I mean they obviously. The the downside of investing in our space is is the sales cycles are pretty long um particularly in gas where you you got to? you got to be very very sure that it's safe before you deploy it. So so there's a lot of extra checks and so yeah, so yeah so so um now they they completely get that the market they get the technology they get the solution I mean they like I mean it's very clear to understand and but but and they just they just have to be able to that to be a little more patient than some investors because of the the long sell cycles. But it's. It's going to be worth it because there's a very big market there and and and the other thing is ah the big barriers to entry I mean it's ah because it's been a very tough thing to develop. We need a lot of investment from us. We need a lot of investment from our customers as well to to get where we are. It's a bit harder for other people. Coming wanting to enter the Market. So so um.
| 51:04.85 | chrissass | So that's it's I always love to hear the business angle usually yohan's asking the business questions. But I think that's pretty interesting. Um I've got as we're pressing it with its time kind of 1 final question from my side is so what surprised you? What did you learn that you.
| 51:12.15 | Adam Kingdon | Um, yeah.
| 51:24.34 | chrissass | Went into this and didn't think you would know so you've been doing this. You've had trials for some period of time. You've got some live customers now. So what's the lesson learned that you did you didn't didn't expect along the way please please share some.
| 51:33.50 | Adam Kingdon | Ah, loads of lessons up. Yeah I can yeah yeah, probably they? Um, ah maybe the maybe the biggest 1 the biggest 1 that that when we started we we thought we. We just need to develop 1 1 system 1 solution gun trial it and then we both start selling it and and we we we then found that that actually a huge number of different variety of governors in the network and and and quite a bunch of them were under the ground as well. And. And our customers explained to us that actually there wasn't any point in being able to do half the governors in a network you had to do all the governors in a network. So for the system solution to be deployable. You need we needed to be able to address these different governor types so that that was a bit of a learning I mean that's basically. Some due diligence. We should have done better at the beginning. but but yeah but we learn a lot um a lot a lot of things I mean it's been quite a quite a journey of discovery the whole the whole last five years. It's miss and we've we've had some really tough problems to overcome and the yeah. But technical teams done amazing job getting over them I mean every time we see we kept coming against a problem and thinking well that that is a real show stoper I don't know how we're gonna resolve that and then after a while you eventually you finally you have solved it. So so um, yeah, we got a.
| 52:48.43 | chrissass | Well.
| 53:03.78 | Adam Kingdon | Yeah, got a fantastic team. We put this product together.
| 53:07.82 | Johan | I Think that's I thought it was really really interesting I when I read up on it in the beginning of of this show I was I was kind of focusing on some kind of sense of date that but I kind of enjoyed discussion and learning about this pressure which I didn't know before. But. Maybe 1 final question and and as as Chris mentioned I used to go into the commercial and the kind of sales approach around it. You mentioned the long sales cycles. Do you see them diminishing. Do you see them getting shorter and shorter because of your own technology and the way you deploy it or is The. The actual sales cycle. The long sales cycle based on the customer's decisions because I think the sales cycle go both ways here and you you now are proven case. The needs are there for the customers. So Why is the sales cycle long or so long as as an obstacle.
| 53:58.70 | Adam Kingdon | Well well I'm sure will shorten and and we we working very hard to to shorten it by making it easier for the customer by making sure all all all ah all the information they need all technical data and everything they need is available them really quickly and easily so they can assess The. Product prop quickly. Ah so we werere doing everything we can and and also we got sgn who we've we've been doing our initial trials with we've been very generous about talking to our new customers sharing information with them and which helps but at the end of the day I think I think. Yeah I think in my experience all utilities have pretty long sales cycles just because the nature of their organization. Um, and gas gas probably particularly because they've got the critical safety and anger as Well. I mean they they cannot deploy anything on their network till they're absolutely 1 hundred percent sure that it's safe and.
| 54:40.72 | Johan | Yeah I second that yeah.
| 54:48.36 | Johan | Are.
| 54:54.96 | Adam Kingdon | And getting to that point sometimes takes time. So so I I think we we'd love to have them sail cycle a lot shorter. But I think if I think they I think we we have to be realistic.
| 55:07.42 | chrissass | So I appreciate you answering that question I think what I've learned Johann through through our conversation today is when when Adam and I first spoke and we were Goingnna talk about sensor in the gas network engaged it. It didn't resonate super well with me had the beginning to be totally transparent and then. As we talk more and he talked about you know the leakitch problem. So ah, the the immediate need started that was interesting that the getting biomethane and and hydrogen and other gases and being able to handle them getting into the the network and and having the same kind of distributed input be different. In a gas network than perhaps it has been in the past that got my attention and then as you move to the kind of data points that that he described with the ai and with you know, getting the forecast and really understanding. And trying to predict what the pressure could be based on. You know what's in the system based on a bunch of factors that would be hard for an individual to do but for a machine relatively easy to do if you have the right programs I think that's that's all pretty cool stuff so you take something that really. Kind of a solid state kind of thing doesn't sound exciting at face value. But when you put all the pieces together like you've done Adam I I think it's pretty cool I think it it makes a difference and I appreciate you sharing the story that to me if I hopefully I've repeated back. What I've learned I hope that's what you were trying to tell us if I'm wrong, please call me out.
| 56:32.37 | Adam Kingdon | Um, you know, very much. So.
| 56:36.42 | chrissass | But um I think it's pretty exciting stuff and and I want to thank you for coming on the show with us.
| 56:39.91 | Adam Kingdon | Well like to thank you very much for inviting me I've I've really enjoyed it. It's been. It's been great. Thank you very much Chris Thanks Very much your high and.
| 56:47.66 | Johan | Great Have you on madam.
| 56:50.22 | chrissass | Well for our audience you've spent another hour listening to insider's guide to energy I hope you've enjoyed the the journey as we've learned about the gas systems and pressure and the gas and I hope you've enjoyed as much as I have if you do enjoy these shows please invite your friends to listen share our link and don't forget to subscribe to both your. Where you get your podcasts and or our Linkedin channel and we look forward to speaking to you again next week thank you very much.

Introduction
The technology
The AI
The business case
Future markets
Lessons learned