The UK Digital Water Utility Experience

I'm going to share with you the suit this morning some of the experiences that we've had in the UK of moving through digital transformation in the UK water industry this is really exciting stuff it's really exciting stuff because in the UK I have seen pictures glimpses of digital solutions really improving asset performance management and customer service I've seen that happen which is great and there's no shortage of technology suppliers who will offer different ways to help you do that which is great but the path to digital transformation of a water utility is not always a smooth one some of the UK water companies started a while ago and they're still not there it's it's a journey that just keeps on going and it has been a little bumpy but I want to share some of those experiences with you so that you can learn from that we can all learn from that and we can get better but before I do that what I want to do is just give you a quick overview of my CV my backgrounds because it will help you understand the reason why I've formed some of the opinions that I have so I started out my career in engineering civil engineering I was actually a wastewater hydraulic Network modeler and I spent eight years doing that type of work supporting the capital program for our for our clients helping them to optimize their capital solutions after eight years of doing that I moved into operations where I spent a couple of years managing a County for seven Trent water seven Trent are one of the largest UK water companies and during that period I was responsible for day-to-day 24/7 management of the assets in that County and keeping the people safe who were doing the the operating and the maintenance of those assets that was a big responsibility after a few years of that I combined that engineering and ops experience and I moved into the strategy department of 7th water in the UK we have investment cycles so my job during that period was to help them with their business plan and part of that business plan was digital transformation how were they going to get smarter and one of the activities that went into that business plan was to create a smart wastewater Network so when my time and strategy came to it to an end I got the job of delivering the project that I had just put into the business plan which was to develop that smartness with wastewater Network so I've learnt a lot about the impact of digital transformation on a business over the years I really have and I think what I what I really want you to remember when I talk about this is that period that I spent in ops because I have been on the receiving end of digital solutions that didn't work and I've been on on you know on the frontline of having to try to get the people who work for me to embrace digital when they did not want to so I know it's not easy but we will go through some of the lessons learnt from the UK today so the digital utility the UK water experience then we'll talk about what a digital utility is so some of the attributes of a digital utility well then talk about how to become digital so I'll take you through a high-level plan of that journey and then I want to take you through some real-life case studies transformations of different parts of a water company business that I've observed so a little bit about the UK for those folks who aren't that familiar with the UK water industry that is a map of England and Wales and Scotland and Ireland the UK water companies serve 60 million people thirty-two privately owned water companies to government run and since privatisation in 1989 they've invested over a hundred and thirty billion pounds in maintaining and improving the asset base so those are those are some high-level figures for you on these maps that you can see on the screen those are the combined water and sewage companies across the UK so those companies look after water production water distribution conveyance systems collection and wastewater treatment and we're seeing a lot of pressure on the on the water companies in the UK pressures that I think you guys can probably relate to we've got similar concerns around aging infrastructure and the resiliency of the assets the level of customer service that we're able to provide there's a general theme being required to do a lot more for less so our next investment cycle is about to begin our next five-year investment cycle is about to start the amount of money to spend in the UK is 50 billion pounds fifty billion pounds total expenditure for the next five years that's exactly the same number as the previous five years but the standards have all gone up in addition to that the third point their common key performance indicators for UK water companies so for the first time now we're seeing the water companies all being judged against the same thing there are seven key performance indicators that include things like sewer flooding pollution from the sewer those sorts of indicators where all of the water companies are being expected to achieve the high standards of the upper quartile water companies and they've got to do that for the same amount of money this was in the general investment plan in the last investment period and to make it even more challenging there are penalties if they don't hit these key performance indicators so where they fail to achieve their targets there's a total penalty over the five-year period of over four billion pounds so challenging times but those are the reasons why some of the reasons why you will see digital in every single UK water company business plan they have got to find ways of operating their current assets more efficiently because they cannot afford to build loads of new ones that's the message here so what is a digital utility what's a digital water utility a digital utility uses data to help their organization and people be more informed and achieve better outcomes data I'd go further than that statement a water company a water utility is actually a decision-making Factory that's what it is whether you're making decisions about which event to respond to which alarm to respond to which alarm to ignore what asset to invest in what type of investments making that asset which customer call to prioritize they're all decisions that's the day-to-day life of a water company and decisions need to be based on data so a digital water utility puts data at the heart of their digital transformation strategy they build data pipelines and automate the collection of data so that it is slick and they integrate that data so that the full business can get use from that data they bridge the gap between all of the vendors specific platforms that they have in place so that they're not constrained to only being able to analyze certain silos of data and that opens up the possibility to start making some much more informed decisions and of course through digital technology you can start to automate that so you generate algorithms and clever rules that suggest suggest an activity whether that's shutter valve or invest in this or respond to that quickly leave that one alone better outcomes come from this and one of the reasons that better outcomes happen is because the system can learn what I saw when I ran operations a long time ago was a lot of historic knowledge in people's heads and the quality of the service that we gave on any given day would depend on who happened to be working that's not great that that made me feel really quite I don't know uncomfortable with this type of system that I'm describing when it suggests an activity when it when it supports a decision if it's correct then the system knows that it it was correct if this if the decision that it was supporting if the answer was wrong then that gets fed back into the system the algorithms improve you don't get that when it's all going on in people's heads so that is key to a digital water utility putting data at the heart of it and of course what you can start to realize then are the massive advantages of working in this way across the full water cycle rather than your waste water network decisions been based on waste water network vendor specific data you can start to join lots and lots of dots data that's related to water production might actually be quite useful in the decisions that you're making about your conveyance system these are the sorts of efficiencies that that's what a data-driven approach can start to bring and it's getting cheaper and easier so this slide is showing the three waves of digital you know this this is what we've seen we've moved in wave one from great we've got computers to wave to where we could start to network and communications improves automation started to become slightly more sophisticated and now we're moving into wave 3 where it's all about big data analytics IOT and and the good news is that is becoming more affordable so again when I look back to when I was doing planning work trying to help my water utility become more digital data storage was really expensive the algorithms were not easy to generate it was a very expensive process to go through that is not the case now storage is cheaper and analytics are a lot more accessible to a water utility so that's good news and of course digital water utilities they're integral to a smart city so results from top 100 hyper-connected smart city studies show those improvements related to water so 40% improvement in public health 37 percent improvement in reliability and resilience 34 percent reduction in pollution and a reduced water usage of 29% that's fantastic what underpins those statistics are things like real time water quality monitoring smart metering and proactive and predictive maintenance you achieve those numbers through knowing what your assets are doing and responding before it results in a failure of service to a customer so moving through how to become a digital utility alright I'm going to go with a caveat now there is no right answer to this question and if anybody tries to tell you that there's some magic blueprints that you just follow and you will be magically transformed that's just not true because every business is at a different stage of maturity they all have their own specific challenges what I'm going to do is just take you through some guiding principles the first message with this fundamentally when you're visioning your digital transformation you need to try to break away from the idea that it's just a collection of Technology projects it is not an IT technology project that's that's the most important lesson I think we start in stage one by doing that visioning piece so what does the water utility want to achieve there's loads of technology out there you don't necessarily have to use all of it there might be some things that aren't really hurting you there are going to be some things that are specific to your organization that you think yeah we really need to hit that no we don't need to worry about that right now but we can put it into the long term so that's where you do your visioning and you start to do your gap analysis I'll show you a little bit more detail on that in a minute but there are certain criteria that you look at within a water utility and assess a level of maturity Jorah for plan of where you want you know how you get to where you want to go and we start to build the use cases this is where the second most important principle comes in people people will be at the heart of a successful digital transformation and that is for two reasons first reason you have got you've got to be able to implement that digital solution it is not good enough to just create stuff that's clever and expect people to want to use it you have got to bring the end user into the development of those solutions and more importantly than that you've got to bring the right end users into the development of that solution so ideally you will select representative that's better representative influential end users so when I again thinking back to my UPS days there were guys and gals who were in my ops team who had some influence get them involved in in the development of your digital solutions that will work much better secondly analytics a lot of what is done is automated but you absolutely have to find ways to incorporate incorporate the domain knowledge into the analytics this is why this is one of the prime reasons that the digital transformation journey has been so bumpy in the UK there was a focus on this being a collection of IT technology projects and all of that knowledge that sat within the business in people's heads was not incorporated into the evaluation and testing of the analytics and again if people are involved in the creation they're more likely to use it so people really important third architecture and data model there is no doubt the technology underpins all of this you need to have the right IT folks ot folks part of that end-to-end team who are going to deliver those different elements of digital transformation there is usually a procurement piece so a water company will have to go out to new vendors to supply some of these different services and the implementation of that so how you start to actually create the services and manage the performance of them so in the same way as your smartphone keeps getting smarter and the apps keep getting better digital transformation in a water utility doesn't stop the customers know their customers are very well informed on what's going on in digital technology they can see it when they pick their phones up so they're going to expect you guys to keep moving so don't expect to just get there and tipton it will take a while so this is a collection of criteria example criteria that we would look at when we're establishing how mature a water company is a water utility is so first of all let's hop how advanced is the strategy envision how much thought has gone into creating that where's it at what sort of level of sponsorship has it got is there a business structure in place to really drive this through the business because it's not easy the utility uses the information to enhance resiliency really the question is how easily is that data being used to improve resiliency is information being used to optimize workforce management and develop the new skill sets that are going to be required over time is information used to manage assets so how well is that data that you're collecting going into a consistent asset management type approach the utility data collects maintains quality and security of data and should transmits it to proper points for analysis so that's around the infrastructure of how it's done systems managing the performance are integrated across the organization so again what we're talking about is breaking down the silos in the water company business and just creating data that anybody can access for whatever reason it might be useful to them and then data analysis methods are used to produce useful and actionable information it's not good enough to just do amazing analysis you've got to have the infrastructure in place so that the business can act on it easily not sit for hours trying to work out what the right spreadsheet is and how to interpret it [Music] so [Music] that was a bit of theory will now move on to some case studies so this is an example from southeast water in the UK where we help them with digital transformation of the water distribution system so southeast water have got very very tough leakage targets as many of the other water companies have they serve a population of 2.2 million people they've already they're already on the journeys being smart they have a network of sensors and the process is in place to do something with the data but it wasn't working and well enough so they looked at moving up that level of maturity specifically they wanted to be able to identify where a leakage repair was required and prioritize it because one leak is worse than another one will have a really bad impact on customers the others won't they also wanted to be able to reduce the cost of detection again through automation and they wanted to reduce spend on repairs so repair the leaks that mattered don't just repair everything straightaway it's just not an efficient way to operate so what did we do well we we worked with southeast water on a program that tested the latest sensor technology we're working with nine different suppliers to look at what sensors gave them the best outcomes in the different scenarios that they were placed we unified the data so remember what I said about vendor specific platforms in this particular activity we unified that data so that all of the data was in one place very powerful and we managed the collaboration between those nine specialist suppliers this was a pilot that we did with southeast water and the learning was well we learned exactly what we hoped we would we learned that there was a better way to do this great we also proved that we could meet those problem statements the prioritize response was faster we could predict events more effectively and we could improve the customer experience because it's really useful for a customer to know whether they're going to be off supply for one hour or 12 analytics can help us do that and we learn a lot of Technology lessons too and this just gives you a little overview of the types of things that I'm talking about so when I'm talking about monitoring of a water distribution network we've got collections of water meters we've got sensors that are managing water quality and pressure we've got acoustic loggers that are out there trying to detect signatures that might indicate leaks the data is transmitted back into the platform that's now taking nine different vendors data and we process that so really nice example of how we took the water utility through that journey I know it's Jonathan there you go so for those who don't know me I'm Jonathan Fitzpatrick I'm the chief inspiration controls engineer for North America and also the North American sector lead for digital solutions so deploying a lot of what we're talking about in the UK in the North American market as well so if we take it back to this side of the pond we have a key stay from Atlanta where they were having a number of similar issues going on the UK a lot of sewer overflows that were affecting the problems with aging infrastructure significant provements and will cap a significant capital program that they need in order to sustain their infrastructure and respond to their population growth and they are looking at ways to reduce their operational costs and they were looking to the digital dual solutions and analytics to do that so the approach here was to find a easy win use case and that was sewer overflows they were spending on average what two hundred thousand dollars every time their sewer at overflow and anywhere from 100 to 200 overflows were occurring in a year in various regions within within Atlanta so it's a very significant cost associated with doing that not to mention the publicity around this when as it's affecting the systems of their city and we have a video from the city manager where he was actually on the news to show how great this this system was or our new technology is saving local taxpayers hundreds of thousands of dollars channel 2's Dave Huddleston reports how a group of city workers came up with the idea that prevents sewage spills look at Crosby it's actually alerting right now Atlanta watershed manager Patrick Woodall is on the phone after receiving an alert that one of the city's 1500 miles of sewer line is clogged have them text me let me know what the results of what they see in 2012 Atlanta experienced the largest sewage spill in its history and was fined more than 100 thousand taxpayer dollars in 2013 Woodall and his staff developed a computer program that tracks all of the city's sewer lines and uses that data to not only monitor levels they can track unusual readings similar to unusual credit card usage we follow Woodall and his crew to this trouble spot of raised manhole in a backyard in Northwest Atlanta they're able to unclog the line and stop a spill before it even starts had that not been in place this would have surcharge you can look around you this is an area where kids or animals pets would have been playing with that computer program and 280 monitors put all in his 9 person team have prevented close to 50 spills including this one so Victor H made something we we start saving money at the same time in Northwest Atlanta Dave Huddleston channel 2 Action News there's some really interesting but takeaways from that and and they actually installed a very large set of sensors associated with us and it really goes to this IOT revolution that we're seeing so the cost per sensor was extremely low totally total installed cost was was about $1,000 a US per site was over $1,000 per site so and that was full that was for the instrument and the installation of labor associated with that and it takes advantage of just the the type of sensor technology that you can integrate by going to an i/o - III ot infrastructure it was tied into the municipal Wi-Fi network so another very low cost way of getting that data back into the system it was also interesting so they that dissent the this analytic that they were running in order to predict where the overflow was going but they were also able to take historical data and run it through the analysis because they actually had some capital spending in place about fifteen million dollars that they had negotiated with the EPA to go in and and to address this issue because they're having problems lower flows in a few of the neighborhoods so when they ran the historical data through the analytics actually showed that the capital program was in the wrong location and that there they actually didn't have a sewer overflow problem in that area and that they needed to needed to address it and they need to spend that money somewhere else so they're able to go back to the EPA and defer that 15 million dollars and focus on a digital approach and focus on upgrades in other areas so it not only did it help them operationally but it also had informed their capital program and and working with EPA EPA was we're pleased that they had the data available to show them that this was this was a viable solution so you know a few takeaways from this the you know there because of the success of this first rollout of their digital solution the the sewer overflow sea of Atlanta is now taking a one step further and they've decided to do a digital transformation project and they're moving towards this one water vision so integrating their systems together a lot of like what Joanna was saying that they're doing in the UK integrating these disparate business units together creating these data Lake layers that will allow them to do the analysis across business lines to improve their operations and and compliance and all these other issues that they're having in the city back to you thank you thanks so as we've been on this journey of helping lots of different clients we've had to help them solve lots of different problems so you can see on screen here is our periodic table of some of the different data analytics that we've created over the years to help our water companies our water utilities meet the different challenges that they're facing so you'll see some in there energy management pressure optimization prediction of events anomaly detection so you know where we want to spot where and where an assets starting to fail those early signs so I'm just going to give you some examples of some of those now before I give you the specific examples I just want to point out that this is the general architecture that we use because all of our clients are in different places along their digital journey we have to be very flexible in the way that we can help them analyze data so this is the solution that we use we use a Microsoft Azure based architecture that allows us we'll move from left to right on the screen allows us to take sensor data from any platform open source as well we always I can't think of an example where we haven't really used weather data as well as other types of open source data as well so in the cloud we can ingest all of that data and we can start to process that data and obviously it's stored in that second client second column there and they're moving over into the third area the analysis section there's any number of approaches that you can use to analyze data from very basic analytics through to more complex machine learning we choose whatever is going to work best for the client and what whatever solution can be incorporated back into their system most effectively and I think this is a really interesting message because everybody can get a little bit caught up in you know the the buzzwords of artificial intelligence and all this you know sort of super super computer you know amazing science-fiction technology levels you know this is what people start to get really worried about but actually you can achieve an awful lot with relatively basic analytics and exit II that you need to as and when required that's that's the learning that we help our clients to go through whilst they're getting their systems and moving through their own digital transformations we can keep the wheels on the bus help them start to meet some of those you know key performance indicator challenges whilst they're getting their enterprise architecture sorted out and we output that in any number of ways so that could be output as a sort of web-based interface where we show the results of the analytics in a dashboard or it can interface back into the the water utility system so that's how we help develop the solutions and this is this is an example so several UK water clients because of the common KPIs several water company clients have got really really tough challenges are wrapped around reducing the amount of sewer flooding that occurs so every water company is expected to reduce the number of incidents where sewer sewage enters people's properties and some of the reduction targets are absolutely you know that they are going to be very very challenging several of the water companies have got to reduce their numbers by more than 50 percent in the next couple of years that's some big numbers and those big penalties associated with not meeting that so how did we help our clients meet that challenge well again we we knew that they needed ability of what was going on and a lot of this flooding from the sewerage Network was occurring quite close to people's you know residential properties we needed visibility of what was going on basically at residential level so we ran a pilot where we are installing 4,000 small sensors across the network at residential level so that the water company can start to gain that visibility of where the sewers are starting to block and the increase there's an increased risk of flooding that's what the sensors are there for four thousand since this is a big number and there's a concern that the control room would be overwhelmed with alarms that were now coming in from these four thousand sensors so they needed analytics to process all of that data and really extract out the alarms that look to be blockage related not the alarms that are just triggering because the the sensors you know struggling with the environment that they're in so the results that we expect from this trial will be an incident reduction and an improvement in customer engagement we've already seen that in in in some of the trials that we've done for the water companies customers love it when you turn up at their property and they say and you say we think there's something about to go wrong don't worry we're here to fix it that is so much better than why we're here where do you want us to start cleaning up its the transforms the way that you interact with your customers and we start to get a better understanding of what's happening across the system where the assets are what the patterns are what wet which areas are we more likely to have these problems in and throughout that whole journey we're demonstrating the value of analytics because it was analytics that shows where those monitors should go in the first place using risk based machine low modeling we selected the locations and now we use analytics to process the data to this is a similar example but a slightly different slant so this is a CSO application so this is where we are helping our clients reduce the number of CSOs again talking to the end users we understood that one of the primary concerns was overload of the control room the water utilities need to put more sensors out they do not need loads of noise in the control room from all of this new sensor technology so the first thing that we developed was an analytic that could verify the data so in this particular example what you'll see we've got a GIS view on our dashboard deliberately nice and simple so anybody can understand it we've got a little GIS view and each of the sense of locations is be color coded according to their status so if this was operating in real time you would see some red which means Oh something Bad's about to happen amber which means you know prepare and green which means everything's fine at this location so you would see that color coding in the GIS view then in the center of the the right-hand panel that is where you can see the number of overflows being counted so each time that overflow activates it increases the spill count in the bottom right hand corner this is our rainfall that we're using as part of the analysis and this little time slider just allows us to move around the time series data and see exactly what's been happening over time in this particular example the telemetry I think that was about five o'clock in the morning in December the telemetry fired in an alarm saying that an overflow is occurring we know that it wasn't the analytics of processed processed that data and that alarm was not real an experienced person in the control room would also have known that that alarm wasn't real but we can automate that and make it more consistent so that's what we've done the telemetry alarm will have been ignored we could with high confidence advised operations that they didn't need to do anything I didn't need to notify anybody they didn't need to respond fantastic and this is an example of an asset that's behaving itself which is great so this one we can see that our spill count is jumping over time it's about a two-week period it's jumping from spill of twenty four up to a spill of twenty nine so five spills five overflows and in the top here the top right hand of that visual that's the that's the basic output of the machine learning that's boiling down the analytics to a really simple how likely is it that that overflow should be operating in a percentage term really easy to understand so each time that spill count jumps up you see a corresponding spike in the prediction the analytics say yes this overflow should be happening you don't need to worry about it this one a different story so in this particular example we've jumped from a spill count of 20 to 23 over that period of time and the model is telling us that the likelihood that a spill should be occurring hasn't exceeded 30 percent in fact it was zero percent at some points so we know that there's something wrong there we know that somebody has got to get out there and respond the advantage with these analytics is not just the real-time operational response though that's a big part of it yes you know you want to get out there and stop the bad thing that's about to happen but the results of this analysis are there for anybody to access going forwards so if you're the operations manager he wants to know what one of those sites is you know what it's been doing this year you can look if you're the person in strategy who's trying to put together a business plan and and deal with the regulator and talk about how often these overflows are occurring you know how many times it's built because you've taken out the false alarms and you know why that's really useful really really useful and then when it comes to your asset management process when you're trying to decide whether a capital investment or an operational investment or no investment is the right thing you've already got a load of analytics that were done in real time that tell you whether the thing should have operated or not not commissioning loads of separate studies very very efficient sentiment analytics okay thanks Jay oh one of the other analytics that we've been looking at is is around customer engagement and around customer sentiment so what what we've done is developed an analytics package which will consume social media data so it ties in with Facebook ties in with Twitter and it will give you the customer engagement score trending topics customer issues and very importantly the location that the issue issues occurring it so what what we've seen actually and a lot of the test data that we've used and we pulled some data from from LA where there's a lot of people on Twitter and you know there'll be a sewer burst it'll be an overflow and they're on Twitter saying hey LADWP like there's a burst at the corner of this and this and that comes in before the before the scale system realizes that there's a burst so we can tie that in with the asset and we can tied in with a region we can show that there are issues we can see issues with billing that might be occurring just general sentiment around the utility and whether it's positive or negative in any given period of time you can see you basically use ever all of the citizens within the city as part of an extension of your SCADA system and pulling that data in so phase one is is what we're we're developing right now which is just taking that sentiment consuming going through natural language learning algorithms and and deploying the issues on the GI Matt looking at who are the influencers in your area right who are the most active people that are contributing to the list the occurrence of issues and then coming up with an overall sentiment score that we can look across utilities and and see generally whether it's positive or negative phase two the plan would be that this actually gets tie back into assets tides back into the SCADA system so you would actually be able to say okay there's our citizens are saying there's a burst at this location we actually think that that's a result of this other analysis that we're doing say using the SCADA data so we can tie that in as another point of information back into the system so this is a an interesting way of interacting with your with your end customer we're finding that there's a lot of emphasis on customer engagement at that local level and it's a great way to see where where we're trending as a utility and to wrap all of this up we are currently actually doing the working with the Water Research Foundation and leading the digital utility of the future framework Roby will be developing you know sets of criteria framework for what a digital utility is there we've got a number of utilities across North America and I think a few international ones as well that are going to be working with us there is still an opportunity if EPCOR and others would would like to join if so just let Joe and I know and we can bring you in but the whole idea is to come up with a common framework based on the feedback that we get from utilities and and people that they serve Joe all right okey-dokey so key takeaways from all this I just want to leave you with some of the messages because I know I've given you quite a lot of information this morning so when you leave here these are the four things that I hope that you'll remember first of all digital transformation is more than a collection of IT technology projects that is crucial involve your people they need to be at the heart of the transformational process they need to be involved in the development of the solutions not just consulted as end-users data-centric architecture will stand the test of time so you might have to come up with strategies where you're moving in parallel and you're implementing solutions now to solve problems now that are maybe limited to certain vendor platforms that's fine but have a long-term vision that is really moving towards having that data like having that data in one place it will serve you well in the long run and then finally digital transformation is a journey I've stood up here as you know somebody from the UK to talk about digital transformation but I don't want any of you to be under the impression that we have absolutely nailed it because we haven't you know that to be honest some of the water companies that went first have had the bumpiest ride there's been so much learning and they are all continuing to try to move to a point of being more digital so how can we help you well this is the way that we help our clients we've got a global delivery water digital water delivery team and those are the types of folks that you would expect to be involved if you are working with us to develop your digital solutions these are the folks that we would be drawing on from from around the world system architects data scientists analytics experts business transformation professionals and operational technologists now you can go and hire in those roles but the advantage that we have at Stantec is that we have already bought that group of people together they all speak their own different languages but over time that group starts to talk its own language they get what each of them means because they've worked together so many times on these types of projects it makes the delivery more efficient and of course we've got a load of water and waste waters domain expertise that we can build into those analytics and this is the type of approach I've already mentioned this but what we will often do with our clients is move through several pilot phases in the areas that are of challenge to them to help move towards that overall digital transformation goal so thank you guys I appreciate pleasure [Applause]

Channel: Stantec

Views: 1,380

Rating: 5 out of 5

Keywords:

Id: V8DEAy3o0S8

Length: 46min 48sec (2808 seconds)

Published: Fri Mar 20 2020