Streamline your pipeline SCADA alarm management for facilities with airSlate SignNow

How to create outlook signature

all SCADA systems are client-server the polling engine is the server right so the client the HMI asks the polling engine for data and the polling engine goes out and gets it if it's a demand request otherwise the server is asking you the polling server is asking for the data and bringing it up and storing it in the HMI so at its very simplest level a all SCADA systems you know their simplest level are what we would call a central master meaning one polling server and one HMI package sitting on top of it could be more than one console but one polling server okay so that would be a central master the next kind of system would be what would be called a distributed system or a peer-to-peer Signet is a distributed system or a peer-to-peer because I can have its polling server in multiple locations throughout my business and any of the HMIS have connectivity can see any one of those polling servers you know provided I have login credentials and capability so cygnets real popular in production facilities for that reason because I can run it in the field and yet glue it together at the enterprise level the next kind of SCADA system is what I would call a master sub master so what what this is and you you don't tend to see these in the u.s. very much they're fairly common internationally where you have big state-owned companies where this would be the control system for a gas utility a city a set of city gates around a community and this would be the interstate transmission pipeline control system and it's polling these guys rather than polling the devices it's polling the servers that are polling the devices so I'm rolling it up so it's a master sub master this approach can be effective and I actually think that it's a it's an approach that's going to make sense going forward one of the challenges we often have in this is I might have a skated implementation at a compressor station but when I go to the main control room the pipeline control center the only thing I know about that compressor station is suction pressure dis prep discharge pressure flow through the station and run status of the compressor I don't have any of the detailed information available to me in the control center if I do a master sub master depending on the tools I pick and how I deploy them I ought to be able to get to all that detail rich data from the control center the other thing is what are we all interested in right now for working for an oil and gas company saving money how are we saving money we're trying to do more with fewer people right so if I've got 24/7 operations here and 24/7 operations here and 24/7 operations here if I can take those two sub masters and make them 8:00 to 5:00 during the day when the workloads high and do everything else from the main control center at night and on the weekends then I can radically shift my operating costs so there's some value in this kind of approach if you can find that the right tools and make it work all right so polling module it's one of the most critical pieces of the system it needs to be able to handle multiple protocols it's got to support multiple media and it needs to support primary polling path and secondary polling paths if the primary path fails it's got to handle Network protocols and encapsulation where I'm taking Modbus and putting it inside of a TCP II and I've got to support backup the real-time data engine its where as I mentioned is I got an image of the real-time data it's where I'll feed other applications that need to see real-time data generally in the real-time database they run that application in the RAM on the server so it's fast right so that it needs to be high performance ideally it's going to be able you're gonna be able to back it up and you're gonna be able to back it up without taking it offline and it should support redundancy as well you also have the historical data engine the big thing about the historical data engine is it's constrained mostly by disk volume how much data I keep so it needs to have algorithms for data reduction so I can set dead bands and you know if for the last six months I want to see data every five minutes but everything that's older than five years I just want to see a daily a daily number right that kind of thing could be redundant most people want them to be redundant as part of the overall system people that operate these pipelines and other systems are controlling they're very interested in the trends the trends are very important to how they keep how they do their analysis and understand what's going on at a detail level we also need alarming so in the alarming one of the new requirements and the regulatory requirements I got to do alarm management to do alarm management I've got to be able to route set priorities I need to acknowledge them I've got to log all that out I need to filter so that only the alarms for this console go to this console and all the alarms for this console go to the other console and I need to be able to get alarms by operating context so if I've got 15 alarms at a site tell me what's the highest priority alarm at that side that kind of thing and the business in if you go to a gas control professional and you say we're going to do a new SCADA system their questions are all going to be about what can I do with the trends that's what they care about they won't be able to build their own they want to be able to save them they want multiple pins all that kind of stuff and as I mentioned that's kind of the business end of a SCADA system for the people that are actually operating assets you also want reports you also need a full audit trail if you're running a piece if you're running a critical system and you have an incident the accident investigators are going to want to see your audit trail what just note to self if I'm in a regulate a control room I do not want to do any logins that are console one console two it needs to be Joe Blow console one Jack Smith console two because the auditors if there's ever an incident they're gonna want to know who was on the desk what were they doing what was their workload how were they qualified and then of course a lot of need for displays in the high performance HMI and there's some good resources out there about this less is more most of the most HMI is built for process control have too much data or the wrong data or the data presented the wrong way so what you really got to do if you're gonna do these HMI as well is you've got to focus on what do they need to operate and there's what do I need to understand what's going on and then what do I need to dig down and do analysis I got I got to keep aware at any given moment is the process I'm responsible for normal or abnormal as matter of form if I click on something I don't want it to pop up in a dialogue and cover something else up because if I do that I could have something alarm or go out of whack and I can't see it I also want to do what's called select before operate which means I want to control this flow controller to a hundred mm add a click you want to control this controller to 100 M add a correct yes click that's select before operate in other words verify before you write there are a lot of cool tools that some of these skata vendors are putting in like pan and zoom and zoom in zoom out and built-in maps and all that stuff nonsense maps are better in a GIS system they're not better in a skater system what a skater controller needs is an overview perspective and they don't really care geographically if it's north or south what they care about is it upstream or downstream so when we build screens everything to the extent we can is left to right like I read a book you do want to do vector graphics and vector texts basically you won't you won't screens to be able to scale well to other platforms so if I have casual users that are using smart devices or whatever I want these screens to size appropriately all right so let me just we'll just take a couple of minutes and see if anybody has any questions you'd like to ask about all this so let me ask was this helpful was this useful okay good good great