Symmetry Flare Workspace Review
Eric Parvin, PE, Parv Consulting
Our company recently completed an effort to import a popular Flare System Analysis software into the new Symmetry process software platform. With years of experience using the first popular software, our client had recently migrated to Symmetry, but hadn’t imported all of the flare system files to it yet. This article is a review of that migration process and other findings in using Symmetry.
As an important note, we did develop a formal procedure for this import exercise just to continue checking that everything works. As you’ll see below, it’s not a seamless or flawless process, but Symmetry makes it extremely easy to move from one flare software platform to theirs.
If you do not read this entire article please take away this one underlying message—VMG service and support is far and away better than anyone else on the market. They are responsive and quick and very helpful.
First, the import process required the original simulation software file to be in the proper format. For more information on this, please contact us directly. For example, when importing from one software in particular, instead of an xml file, the original simulation needed to be in a fnwxml file.
Upon the import for a small system (e.g. 5-10 PSVs or blowdown valves) the import process was extremely easy and flawless. Compositions (including hypotheticals) migrated over, as well as the piping details. Scenarios were ported over accurately as well.
For a larger PSV system (e.g. 50 devices to a flare header system) we did find one pipe was imported but not connected to the flare—requiring literally less than 5 minutes of “repair” work. This was reported to VMG and they looked into the issue and believe they have resolved this issue for this particular flare system and future conversions.
On a very large flare system (> 250 PSVs) there were similar instances of pipes not connecting, but we didn't go through and "repair" it as this was an experiment only, but it was impressive how well it moved that many PSVs over along with all the many scenarios.
We found that the process of choosing scenarios to evaluate with a check box system was extremely easy. Other simulations for flares typically are more difficult to go through and select to run. The graphical user interface (GUI) for reviewing what PSVs were active and ignored was extremely easy to review and it simultaneously showed potential errors on the screen. We could make out which PSVs were active and which ones had problems on the same screen, which usually in older software is not as easy to review.
While reviewing the import of the relief rate requirements from the old software to Symmetry, we noticed that the GUI for programming the PSVs was actually very user-friendly. For example, in the older software, you can input the PSV size and determine relief rates with PSV Kd values and such, but with Symmetry, it’s much easier to input and verify results. Additionally, you can choose both API PSVs and ASME orifice areas (not just API).
One drawback that we gave feedback on is that when choosing an ASME PSV size, the letter designation isn’t necessarily consistent with the area of the orifice (whereas with API, it is consistent). VMG is working on avoiding confusion on this in the future (and may already have a fix).
In the older software, you could manually enter the relief rate required and rated capacity but Symmetry does a better job of checking your inputs and calculating the rated capacity of the PSV by inputting the correct PSV details (e.g. Kd values, Kb, etc.). However, in unique situations, this may cause issues in back checking work and the engineer may have to manually modify the Kd value to get the rated capacity of the valve to match, whereas in the older competitor software, it was just a simple entry. From a pure engineering and QA/QC perspective, I personally much prefer the Symmetry entry method and checking results.
The only place I can see a problem with this method in Symmetry is supercritical mass flux equations for rated capacity vs. required capacities, and possibly 2 phase relief cases, the rated capacity of a PSV calculated by external software (e.g. iPRSM or PSPPM) may not match exactly without the user modifying the Kd values or other coefficients of the PSV sizing equation to force it to match. This has not been fully vetted by our company with Symmetry, but it’s being reviewed for future efforts. Moreover, you could have a PSV with 2 relief scenarios—one supercritical the other not. Having the ability to manually input rated capacity in these special situations was also feedback sent to VMG and they are working on providing allowance for that input in Symmetry, or a secondary consistency check on the input screen.
VMG’s response to the above: "At this point we do all sizing/rating with HEM method. This method is recommended in API for two-phase flow, since it follows thermodynamics principles and integrates flow along the isentropic path. It can be considered rigorous and gives continuous results through fluid regions."
Notice a theme here—they receive feedback and respond to it extremely well.
Other features we were pleased to find:
- Consistency checks—Symmetry flags for attention when there’s an inconsistency. This was flagged in a few cases where the old software did not flag this for attention.
- Speed of calculations appeared to be faster. On the same computer on a small flare system, there was a difference. We have not done a side-by-side test on a large (>50 PSV) system, but it was noticeably faster.
- Symmetry flags for attention “errors” in the flare results, for example high back pressure, and specifically calls it out for the user to pay attention to. In the older software, it was subject to seeing red numbers on the screen and then scrolling through menus to display numbers and interpret the results, an obvious time consuming interpretation of results hurdle.
- Isentropic calculations are easy to flag for calculation in both the old software and the new Symmetry.
Here are some examples of a powerful tool we have found in Symmetry:
1. If you have a global fire case with multiple PSVs relieving—in the traditional software on the market, it was extremely difficult to dynamically model a fire case, and we’ve only seen a presentation at conferences once or twice of someone successfully doing this—and the reports were that even with experienced people in those other software platforms, it was cumbersome to run as well as translate those results. With Symmetry you can actually program the vessel sizes, and schedule a fire event (in Dynamics mode) and model the dynamic behavior of the flare system. Therefore, if you have 10 vessels in a fire zone, you can model how quickly they will reach overpressure and then begin to vent. The results then give you a much more accurate assessment of back pressure on PSVs for fire zones. Traditional software in steady state mode will assume all PSVs relieve to the flare header simultaneously—which can cause significant back pressure reporting issues that may never occur in the flare system during a real fire event. The savings that a company could incur by evaluating this dynamically can be extremely valuable (especially for midstream and upstream companies), but even more so for refinery companies for not only fire cases, but total power failure, partial power failure, and other similar “non fire events”. We have done this for a refinery client in the past using hand calculations and timing, but my thoughts are “Finally a dynamic flare modeling solution has come to market that most people can use with some minimal training.”
Other examples of where this is useful:
- Global (e.g. power failure) cases with PSVs opening and closing at intervals
- A blowdown case in a large flare system may show a high back pressure in steady state mode however, the flow to the flare tip almost never reaches the blowdown instantaneously. We have actual operating data from a refinery client where the flow to the flare tip “peaked” at about 70% of the maximum blowdown valve rate. This can now be modeled in Symmetry without having to rebuild the entirety of the flare system in separate dynamic modeling software.
2. The fittings of the pipes in Symmetry can be sorted and evaluated in order of the fluid experiencing the fittings. In the older software, this was not possible. For example, in the old software for the same pipe size, you simply input the length of line and the quantity and types of fittings, with no (easy) ability to program the fittings in any particular order. In Symmetry, you can program in more accurate isometric details. This is especially important in evaluation of the potentially high velocities (e.g. >0.9 Mach) as where it occurs in the isometric details may impact whether there’s a high sound power level (SPL) issue, or no concern at all. If there is a high SPL, then the Mechanical Engineers have better information for a finite element analysis if ever needed to evaluate. This is more important in some of the cases we’ve run with clients than the simple input format of older software.
3. The overall user interface was extremely easy to learn. Without any formal training, but just about 30 minutes on the phone with the VMG Support team (due to an aggressive schedule) we were able to get things imported and figure things out quickly and easily. We had a few questions but the Support team at VMG was far superior to any other experience with any other software vendor we have ever experienced in over 20 years and multiple software providers. Even though we have working experience with Symmetry and the older "classic" style of VMGSim, we will be attending the trainings in Denver just to "sharpen" our knowledge of the software. I invite folks to attend as well, it's free and if you're in the area, you want to attend this.
4. Originally a problem for us—the flare tip velocity was NOT being reported with the given open flare tip size. We were able to add this simply by programming a line size on the outlet stream leaving the flare tip. However, after our feedback was provided to VMG they have since added a flare tip velocity calculation block to the flare tip within Symmetry and have fixed this problem in just a short month. New rollouts of this feature are available now.
Now for the problems—we’ve told VMG and they are addressing these issues and improvements are coming.
1. We found that on simulations where the K-fitting value of the flare tip was imported as a zero value AND the flare tip open area and pressure drop vs. flow curve were provided, the simulation would run and show a sudden jump in velocity across the flare tip or flare stack. This was an error. The fix is easy—delete the “zero” value in the K-fitting value and it works fine. Update: VMG has already fixed this in the latest release version.
2. Supercritical cases—user “beware”—it’s unclear whether Symmetry equations for rated capacity when inputting the PSV orifice / open area data is using mass flux equations or vapor sizing equations. When you have supercritical relief situations, Symmetry may (or in newer versions may not) over-estimate the rated capacity of the PSV. We were able to validate liquid and vapor (below supercritical) equations and get a consistent match.
VMG response: "VMG consistently uses mass flux (HEM) equations to solve sizing/rating, regardless of fluid phase. The part that varies with phase is the Kd used. The user has a variety of options for this: directly specified value per phase, weighted by volume fraction, or to use vapor coefficient when the valve is choked (regardless of phase)."
3. Defining the atmospheric site pressure in both the original software and Symmetry will be critical to get consistent results. In Symmetry, just poking around won’t reveal it, but it’s an easy solution—just right click anywhere on the main PFD, choose the top menu option shown, and you’ll get to the site atmospheric condition.
4. Symmetry will report errors for high back pressure for non-opening PSVs—for example, if there’s a total back pressure on the flare system of 10 psig, and there’s a non-opening PSV (it remains closed during the relief event) with a set pressure of 75 psig, then Symmetry will report this as a high back pressure error. Update: VMG has added this "ignore" option to the latest release version.
5. In the current version of Symmetry, the pressure reported is unclear whether is static pressure or total pressure. Update: VMG is making the following pressures available in reporting in the next version(s) being published soon: static, dynamic, stagnation pressure.
All in all this was a good experience and I can say with certainty that by using Symmetry, we can:
- Program isometric details with more certainty.
- More quickly interpret hydraulic results and identify “errors” in the flare system.
- Bid on engineering evaluation of flare headers at a slightly lower cost than traditional software (ball park estimates are in the 5%-10% range).
- Produce overall results much more easily—this is very significant finding for us versus in the older software where we had to manually print PFDs and show the issues.
- Perform dynamic evaluations of flares using the original flare file and not the segregated dynamic software, as other software suites require.
The value of the dynamics for most large flare systems (e.g. refineries, chemical plants, some larger cryo plants) can be invaluable and easily pay for the cost of the dynamic modeling software. We’ve seen case studies in the past where this saved >$1MM by not having to replace flare headers entirely through sections of the plant. Please remember, when gas goes from a PSV tail pipe and into a large flare system—it will still flow both ways until the flare header is “packed” with all that pressure. We see it all the time with operating data. Having the ability to model this now is extremely powerful both from an operations standpoint (avoided capital projects) and an engineering company sales point.
Eric Parvin, PE
Please contact your local VMG office for more information about Symmetry.