Synergi Gas Time Varying Module
DNV GL’s Synergi Gas Time Varying Module (TVM) enables you to perform a series of consecutive steady-state analyses to simulate the changes in your network over time. In this blog post, Alex Hollis provides a detailed look at this modeling tool and illustrates the wide-ranging uses of TVM models. Alex, who is based in our Mechanicsburg, Pennsylvania office, is a Senior Principal Product Engineer for DNV GL’s Pipeline Solutions Portfolio. He is a 30+ year veteran of our team who shares his expertise, knowledge, and tips for leveraging Synergi Gas features and capabilities.
The best modeling tool you’ve likely never used
In the world of offline network analysis applications* for both liquid- and gas-system modeling, a great many studies are done using steady-state analyses. Here we are going to talk about using TVM for steady-state gas system analyses. It’s a Synergi Gas Module that does what is commonly called “extended period simulations.”
Putting it simply, TVM runs consecutive steady-state analyses at user-specified time intervals over a user-specified time period. Let’s look at a couple of examples.
TVM Example 1 – An underground storage (UGS) system
This is a sample model of a “salt cavern storage” system. It is a generic sample model we concocted and thus does not represent any real systems that might exist. On the left side, we have active reservoirs.
The grey piping and facilities on the right side are six future reservoirs. They are initially disabled in our model, but they will be enabled at various times during our simulation.
Gas is injected and withdrawn from the reservoirs seasonally. We have compression configured to withdraw from and inject into each reservoir, as we see in Figure 2.
Our modeling objective is simple. In our design, we’ve proposed a withdrawal and injection schedule for the reservoirs, assuming a constant pressure at the “connection point”, we see it in the bottom center of Figure 1.
Our model will run for a three-year period, and we’ll evaluate our planned withdrawal and injection schedule. It may need some modifications. Once we’ve settled on a good operational schedule, we would (in a real-world scenario) start sizing valves, compressors, and so on. For this blog post, we’ll just look at the schedule.
Each of these reservoirs are caverns of known size – and thus volume – and during the three years, we’ll see the reservoir pressures decline during withdrawals and then increase when they are refilled. We will not go into a lot of detail here, showing the east reservoirs as they “go online” during the simulation, rather, we’ll just look at one reservoir to show you the results. Is our proposed schedule for withdrawal and injection feasible?
TVM is set to calculate the steady-states on one-day increments. So, at the end of the 1,155 days of analysis, we have TVM output as shown in Figure 3.
There’s a lot going on here, and we won’t go into details about how we set up the axes for this chart or which variables are which.
Instead, we’ll make a couple of observations:
- TVM lets you model long periods of time; here we ran from 1 November 2020 to 31 December 2023. That’s 1,155 days – and since I specified a one-day timestep, that’s 1,155 continuous solutions.
- The keyword there is continuous. A time-varying model maintains continuity between steady states. This means that as we go through time, Synergi Gas calculates the flow out of the reservoir and thus the volume of gas removed. On the next balance, the reservoir pressure is reduced as a function of the volume removed in the previous solution. While these are steady-state analyses, Synergi Gas TVM integrates the flow rates to determine volumes and uses those data for the boundaries to the next solution.
TVM Example 2 – A natural gas distribution system
TVM can be a pretty handy time-saver for planning studies for gas distribution systems, too.
Let’s take the model we see here in Figure 4, a typical mid-sized city’s multiple pressure level distribution system.
We have current data as to the peak hour flow rate by month (Figure 5).
Our hypothetical company’s hypothetical marketing department has provided some very, very highly precise forecasted demands for the next few years:
We’re expecting the system demands to increase by five percent every year.
The purpose for analysis, then, is simple – can our system provide those forecasted flows, or do we need to reinforce them? And if we do, where? In other words, we are pretty much doing the standard things you do all the time. In a typical model, we’d probably run a couple of analyses for every year, a Winter Peak day and a Summer Peak day. To set them up, we would load the system for the winter day, set the ground temperatures so that our heat transfer calculations will be included, solve the model, and check the results. Change to summer conditions, solve, check the results. Take the first winter model, increase the demands by five percent for next year’s case, solve the model, and check the results.
Then summer. Next winter. Next summer. When it’s all done, we’ll have ten sets of analysis results to review.
Or, we can do it all in one TVM analysis. Yay, us!
To set it up, I created a profile, or curve, of each month’s peak hourly demand and applied it to all our system’s demand nodes. Then added a modifying curve to that curve, one that multiplies the flows by five percent over the next few years. Finally, I added another profile of the average ground temperature for our system.
Our model will start running on 1 January 2020 and will run to 1 January 2025. TVM needs a specified timestep – I chose five days. I could have chosen something different, maybe every 90 days for quarterly analyses, or every one day for real precision, whatever makes sense to you.** The reason I chose five days was really complex: 365 divided by five yields an integer number of time steps. I could have just as easily chosen 182.5 days or 91.25, or whatever.
And then I started TVM. For some perspective here, we’re running for 1,825 days, at a five-day time step, which means that TVM will be solving this model 365 times. In other words, I’m doing all those “ten sets of analyses” at once, and instead of having ten sets of output to review, I’ll only have one.
On the laptop that I used, the analysis took 19 minutes and 11 seconds to run. And I spent about three hours setting the model up for the analyses. In short, maybe 3.5 hours to do a five-year study.
And the best part of it is that when it’s done, we have a lot of options as to how we’d like to get the results out of the model:
In short, you have a huge reserve of very detailed results available to you – all the things you have from any steady-state analysis, except here they’re presented through time. When you run a TVM analysis, we use something called a “Results Plan.” The Results Plan is a list of variables – node data, pipe data, regulator data, compositional information, temperatures, compressor results, whatever – in which you’re interested.
We save them to what is called a “Store File.” And those data will always be available. Let’s say you forgot to generate a list of flows through regulator station BR-549, but it’s an important (possibly undersized?) regulator. You will not need to run the analysis again. Just re-open that model, and Synergi Gas will automatically re-open the storage file. And you can create the report for that station then and there. Save the storage file, somewhere, and it will always be available. With TVM, you also get a storage reader utility, and you can actually pull variable data from several storage files (analyses from this year v. next year v. five years ago, as one example), and do all the reports or charts or lists of those data.
More Examples – Other uses for TVM
We’ve only covered a couple of uses for TVM, but there are many others we could mention –
- Long term analysis of gas gathering systems to model well decline through the life of the field.
- Growth plans for any system – gathering, transmission, or distribution systems.
- System tuning – SCADA data over a period of time can be provided to TVM, and the Model Tuning tool helps you determine system efficiencies or pipe roughnesses over time.
- Transmission system seasonal modeling and bidirectional system analyses.
And so on. Be as creative as you wish; it’s an enormously flexible modeling tool.
We are delighted to answer any questions you might have or to discuss potential applications with you. We can provide a temporary license for evaluations and can prepare demonstration files for your use. So please feel free to contact us through your DNV GL Business Development Manager, our customer support toll-free number at 1-800-800-7764, or you can contact me directly at 1-717-724-2935.
*Where offline generally means studies that are not dependent on real-time SCADA data – the kind of analyses done for expansion studies, greenfield systems, distribution system models, long-range planning, and so on.
**Synergi Gas lets you define your own units for everything, just give us a conversion factor. So, you could run time in Fortnights if that suits you (fun fact, a fortnight is two weeks, that is, 14 days). The sensible-to-use unit of time is “Month,” of course, but it’s pretty much impossible, because thirty days hath September, April, June, and November; all the rest have thirty-one, excepting February alone, and that has twenty-eight days clear, and twenty-nine in each leap year.