The age of the industrial Internet is upon us, which means water and wastewater utilities are now taking advantage of newer technologies that are changing the landscape of our sector. With a few pieces of software, utilities can now optimize their operational performance, enhance asset reliability and accurately monitor their systems.
Some of the brightest minds on the planet are developing software for the next-generation of Industrial Internet technologies, and the potential advantages for the industry are nearly unlimited. These products allow for more intelligent management of assets by harvesting and processing massive amounts of data to produce actionable information.
The rate and complexity of this technological development can be both exciting and unnerving, even it may be unclear how everything fits together or what to do with all the available information.
So how does one even begin to incorporate it all?
Adapting to the technology can be tricky, but the first step is mentally preparing for this fundamental shift in your operational philosophy.
Getting started can be the hardest part, but it’s also the most important step in protecting your operation from unplanned downtime that can directly impact the bottom line; when optimized, these technological changes could save your organization millions of dollars.
It’s better to start with smaller changes, especially for your employees – this allows them to gradually get comfortable with new processes before adopting more advanced technologies.
Most importantly, a well-formulated plan will help your organization transform from a reactive and unpredictable operator into an efficient, intelligent and highly-profitable one where data is now your best business partner.
A Modern Utility’s Guide to Digital Efficiency
Free Event, hosted by the Metropolitan Sewer District of Greater Cincinnati, the City of Akron, and the Akron Global Water Alliance
Up to 7 CEUs Available
Nov. 12 – Cincinnati
Nov. 18 – Akron
Monitor Your Assets
To kick off your journey, let’s begin with basic monitoring of your most critical assets.
This is done by equipping critical machinery with sensors and control networks that read and manage the actions of that specific piece of equipment. By collecting and simplifying the data produced, an operation starts to form a solid foundation that determines how all the separate parts of the system are supposed to function as a complete unit.
While experienced personnel might believe they are very good at catching and solving problems before they become much worse, data is much more accurate and more reliable – it never takes a vacation or sick day. In other words, data is the key to all the dirty little secrets your machines keep from you.
As the industrial process unfolds, the installed sensors pick up information about the speed of each step, the volume of liquids, the internal temperature of a pipe, and where a leak may be letting air or water into the system.
The sensors and controls are doing the heavy lifting for you, which eliminates any doubt about what is happening within the machines. Once the data is harvested and funneled to one location, engineers can now watch and focus primarily on the data from an entire operation; they can access the information in a central office or remotely.
One way to access the system data is through an online Human Machine Interface/ Supervisory Control and Data Acquisition (HMI/SCADA) system at the point of control. This technology allows operators to view all systems in the process from one location, giving them greater access to real-time information and allowing them to make more informed decisions.
The other way to evaluate the data is through the cloud, which allows access from an established control center, but it also enables engineers to examine real-time data from a wireless Internet connection; this could be within the control center or out in the field. Now, any data being collected on the sensors can be analyzed at a central location, and controllers can remotely manage their assets.
Mobilize Your Workforce
Let’s take monitoring a step, or many miles, further. Using mobile applications, engineers have front line access at their fingertips from anywhere they can connect to the web.
Engineers can log in using apps on their iPad, pull up data for all systems under their watch, and monitor real-time information. Insightful and actionable information is always at the ready, so informed decisions can be made without delay, which helps avoid disastrous results and reduce unexpected issues.
Regular maintenance and upkeep will still be required, but the technology definitely helps keep small surprises from becoming major disruptions. Think about the cost avoidance from preventing a water line break: those costs alone makes it well worth it to know ahead of time when things might be starting to break down.
Fixing problems based on data and generally catching problems before they become catastrophic is beneficial on its own, but it still requires constant attention; even manual control leaves a lot of room for error. If that worries you, then you’re ready for some advanced analytics to take you to the next level of protection.
Diagnose Your Pain Points
When analytics come into play, the benefits of intelligent data come into focus – this is when your organization begins to learn the answer to “why?”
By deploying an effective analytics strategy, you can begin to understand the data at a deeper level.
Even if history wasn’t your favorite school subject, you’ll learn to love it now because your data history holds the key to preventing major disruptions and optimizing your systems in the future. Instead of physically mobilizing the maintenance team, trained personnel can check the history of the data to discover the root of the problem. Once identified, the team will be able to write a new analytic to give warnings next time a similar data trend begins.
Predictive software works in a few ways. Some systems use cluster analysis or similarity-based modeling. Others use Proportional Integral Derivative (PID) loops to save millions of dollars in unneeded equipment wear-and-tear that can result from large-scale process variability. These systems identify normal equipment behavior as they predict future behavior. This level of intelligence takes numerous variables into account, so these highly-complex relationships (with the data that accurately predicts future states) are based on any particular variation in one or more inputs.
Here’s how it all happens.
Imagine you’re the only person who drives a given road every day to work – there’s no traffic, no lights, no turns.
This is how a machine would work if no other factors came into play, but it’s never that simple. As the area around your commute gets developed, you start to encounter stop signs, crosswalks, construction that blocks a lane or two, and (usually when you’re running late) you get stuck behind a garbage truck. Even though every day your drive is different than the last, you gradually learn from experience, so you’ll know how to react to each future interruption for a safer commute.
This is essentially what happens with advanced analytic modeling. New input variation – new equipment, new materials, new processes, temperature and seasonal changes, and so on – has an effect on the function of each dependent system. Engineers can trust that a certain fluctuation of one input will have a standard reaction in a system down the line.
For example, a drilling operation in a harsh environment with volatile weather patterns would have an established analytic model that determines what the proper internal temperature of a machine should be, compared to the outside temperature. This allows for variation in “proper” internal temperature and makes it easy to distinguish when any deviation is acceptable or if it requires further examination.
The frozen winter months in Michigan will keep internal temperatures on the low side, but it’s a different scenario when everything starts to thaw: you will know how strong the correlating rise in machine temperature is supposed to be, or if the deviation is too extreme and signifies a real threat to the system’s stability.
Learn from History
They say those who ignore history are doomed to repeat it, so why limit how much history you can reference?
Water and Waterwater treatment both produce a massive amount of data that is eternally useful; limiting your amount of data only limits your ability to make informed decisions. Industrial Big Data gives you the freedom to mobilize much more information than previously possible.
With the right applications, you will be able to gather data from more sources over a greater duration of time than ever before. You can compare all this data to your real-time activities to make quicker, more informed decisions than ever imagined.
In addition to the emerging technology, cloud-based applications are a major factor with Industrial Big Data because massively scalable storage environments can collect from (and disperse to) variable locations give you the flexibility to work on more data and do it faster.
For example, solutions built on Hadoop®, can handle large data sets by clustering large numbers of low-cost commodity computers together to act as a single storage entity. Historian software allows for economical storage while maintaining rapid access speeds. These solutions scale vertically, for any velocity, as well as horizontally, for any volume and variety.
Optimize Your Operation
Now that your operation is working smarter and leaner, and you’ve built your trust in predictive analytics, you’ll be ready to move further down the efficiency continuum. More sophisticated software analytics won’t only catch problems before they happen, but they also allow you to standardize your process across all parts of the operation to optimize performance and value. Electronic instructions are provided to operators to ensure that the best methods of work are followed every time at every place, regardless of experience or training.
The most modern predictive analytics software available now allows you to move from self-defined analytics to a stage where machines learn on their own and constantly write their own analytic rules. You could call it artificial intelligence, but it’s a very realistic shift from unplanned to planned downtime, which could potentially save your company millions of dollars. Predictive-analytic software reduces maintenance costs and improves asset lifetimes to a greater degree than human monitoring could ever do.
Yes, we will still require human interaction and maintenance, but some things simply can’t be done manually. Advanced software can adapt to variations much faster than humans, where the system sends signals to automated controllers that keep everything running in optimal balance.
The point of control goes from the HMI/SCADA system (or iPad app), and becomes starts to become embedded into the controls hardware itself. The information runs in a closed-loop system that greatly reduces the human element, requiring manual interference only when a major issue arises.
Now Dive In!
When all layers of software are incorporated into your maintenance and oversight strategy, you have reached your destination: you are a lean, mean, data-using machine. By taking each step one at a time and steadily marching towards the goal of automation, your oil and gas facility will be at peak performance in no time. Without having to worry about managing emergencies or wasting money on unnecessary procedures, think how much better your job will be at a safer, efficient and more cost-effective operation.
About the Author
Alan is the Vice President of Water/Wastewater at Gray Matter Systems. Alan has over 20 years experience working in water/wastewater and is an expert is water distribution information technologies. Alan is a frequent speaker at industry events around the country.
This content appeared originally as part of “Damned If We Don’t!: Ideas for Accelerating Change Around Water” and is being reproduced with the permission of its author, Alan Hinchman.