Venting excess pressure from buildings protected by inert gas or clean agent suppression systems is critical, but the data used to calculate the required vent area can sometimes be flawed, as Paul Coxon explains.

Incorrect pressure venting calculations can lead to severe damage to buildings

Abstract: For many years the use of gaseous fire suppression systems to protect high value or business critical risks have used calculations to determine the required venting areas, to prevent protected risks from being exposed to pressures in excess of their design strength. The calculations used for determining the required free vent area (FVA) have either had a default figure for the performance characteristics or co-efficient of the pressure vent to be used, or this has been ignored or misunderstood by manufacturers.

This article looks at a series of live gas discharge tests carried out using IG55 (inert gas) and FM200 (clean agent) and compares the calculation system used (VdS Software) of FVA. It looks at and introduces the use of a dynamic co-efficient or resistance to opening pressure factors (RTOP) for the actual performance criteria of three types of pressure vents and their direct effects on the risk peak pressure achieved.

Due to the critical nature of venting excess pressure from risks protected by inert gases or clean agent systems, we explore the key elements required to accurately calculate the pressure venting needed, which includes the required free vent area and the dynamic co-efficient or RTOP factor to provide a true and safe application of pressure venting systems.

Introduction
Gaseous fire suppression systems have been around for many years and the system components, design software, designers, installers and risk integrity are controlled and detailed in British and international standards. Approvals have been developed to ensure that these systems are designed and installed correctly and provide a safe and effective fire prevention tool for commerce and business.

Pressure venting is possibly the only area that has not kept up with the development of standards and is one of the most critical. We still have systems being installed using calculations for the free vent area (FVA) only, and no other information is fed into the calculation for the performance of the type of pressure vent selected for the system.

The German standards and testing body, VdS, provides an approved software system for calculating gas flow rates and system performance and this software has a default pressure vent co-efficient of 2, which can be changed to accurately predict the required vent size. But in my experience, very few fire protection companies change this to reflect the actual performance of the vents used.

In this article, we look at the use of a co-efficient and demonstrate the importance of using it to provide an improvement in system performance. We also look at structural design strengths and correlate the effects of faulty venting calculations with structural strengths.

Structural strengths
Firstly we need to understand general structural strengths of buildings. We have evaluated data from a number of papers on the effects of blasting on structures,^{1, 2,} to earthquake hazards,³  to pressure dynamics of clean agents⁴ . There is limited data and studies into this subject.

What pressures can we work to and what peak pressure can be created?

Example 1: From AFP Report on Pressure Venting, verified by BRE ⁵

The above graph shows a discharge test using FM200 (42 bar system) in a risk of 44m³ volume. The risk was essentially sealed with a leakage of 0.0089m². The FM200 retention time estimated by a room integrity test was 16.7 minutes. The same tests were carried out with IG55 which had a retention time of 46.5 minutes. This gives a good indication on the hold time verses peak risk pressure and should result in a failed integrity test as the risk is too tightly sealed. The risk pressure achieved was negative of around -500 to -600 Pascals pressure (pressure sensor stopped at -189 PA but by analysing the data we can ascertain the pressures) and the positive pressure reached 1044 Pascals.

Example 2: From AFP Report on Pressure Venting, verified by BRE ⁵.

The graph below shows a like-for-like test carried out in the same enclosure as the above FM200 test but using IG55 (inert gas system). It gives an actual peak pressure that was achieved in a risk with 0.0089m² of leakage which gave a retention time of 46.5 minutes with no pressure venting.

The pressure reached was 3000 Pascals as shown below:
 

The above gives an indication of peak pressures in discharge scenarios but based on Boyle’s Law, the maximum peak pressures can be 55,000 Pascals for inert gases and for FM200 8,000 Pascals. With these pressures in mind, we need to be aware of the maximum structural limits on modern building systems.

Watch video of the pressure tests:

In Mark L Robin’s paper on Pressure Dynamics of Clean Agents ⁴ exterior wall strength is detailed in his Table 2. The structures tested were 10 feet tall wall structures. The table below has taken these figures and converted them from pounds per square foot (PSF) to Pascals.

These limits do not take into account joints or tie in to other elements such as floors or ceilings and should not be used as general structure strengths.

Also it is not clear from the report what type of dynamic pressure was used, as a blast pressure will have a completely different effect to an increase of pressure over 5 to 10 seconds on a structure. These pressures appear very high in my opinion.

Table 1 Strength of Walls. Exterior Walls, 10 Feet Tall

The maximum allowable pressures shown above suggest that in a number of cases, there is not a problem with chemical clean agents such as FM200, which produce peak pressures of 1500 Pascals. But if we take into account structure joints, these maximum allowable pressures drop significantly.

To illustrate this, there has been a number of structural failures with all agents but in 2008, a number of composite steel sandwich panels rated at 1000 Pa were blown out by an FM200 discharge. In other examples, block walls rated at 700 Pa have cracked and moved with IG541.

For example in Fragility Assessment of Light-Frame Wood Construction Subjected to Wind and Earthquake Hazard,³ roof panels are rated at 1.6psf or 76.6 Pascals with roof to wall connection at 717 Pascals.

So it is easy to look at individual strengths of structure materials and not account for joints and fixing processes, as well as test data created using different types of pressure dynamics applied.

With the potential of clean agent inert gas systems and chemical agent systems producing peak pressures of around 55,000 Pascals for inert gases (300Bar) and 8000 Pascals (42 Bar FM200) for chemical agents, the component of correct venting calculations is critical.

The industry standard maximum allowable peak pressures of 500 Pascals for heavy grade construction, 250 Pascals for medium, and 150 Pascals for light weight construction provide a safe limit to work with.

Dynamic co-efficient – resistance to opening pressure
In tests carried out in April 2008 by AFP Air Technologies and witnessed by BRE ⁵, a series of live gas discharges were carried out using IG55 inert gas systems and FM200 chemical clean agent. These tests were designed to test the performance of three types of pressure vents in live discharge operation.

Both FM200 and IG55 systems were designed by LPG Fire Ltd using their VdS calculation software for the inert agent and their LPG chemical agent software for the FM200. The risk strength was calculated at the maximum structure strength of 500 Pascals which resulted in a free vent area for IG55 of 0.083m².

They then carried out a discharge test using an open hole of 0.09m² which equalled the maximum opening of each pressure vent tested. The resulting peak pressure was 219 Pascals. This was based on the default VdS vent co-efficient of 2.

The SHX type vent during the tests

AFP then carried out a test on their old HXD pressure vent which provided a free vent area (FVA) of 0.072m² at the achieved peak pressure of 349 Pascals. This was 80% of the required FVA of 0.09m². A further test was carried out with their current SHX pressure vent which provided a FVA of 0.076m² at the peak pressure achieved of 251 Pascals. This was 85.5% of the required FVA of 0.09m². The final IG55 test was carried out on an industry standard pressure vent which was a top hinged bottom weighted (Gravity Vent) design still current today. This vent provided a free vent area of 0.0274m² at the peak pressure achieved of 790 Pascals. This was 30.4% of the required 0.09m². The VDS calculation was based on a peak pressure of no greater than 500 Pascals.

Each of the pressure vents tested was stated (by the manufacturers) as providing 85% to 90% free vent area but in testing they produced very different results.

The reason for the difference in peak pressure is simply the difference in the dynamic co-efficient or as we like to call it, its “resistance to opening pressure” (RTOP) factor. This is the force exerted on the blades of the pressure vent when exposed to a very rapid increase in pressure or blast and the resistant back pressure produced. This is unlike a co-efficient used for fire dampers which is a fixed test with the damper open and a set air flow with the pressure drop measured, which provides the result.

With a dynamic co-efficient, the blast or rapid increase in pressure and resultant air flow, as well as the position of the vent blades, is measured over time. This will be measured in the space of one second. The only true way of achieving this figure is to carry out a live test which must be based on the VdS Calculation software and the type of agent or a pressure simulation of the discharge dynamics of the agent.

In the result from the ‘top hinged bottom weighted’ vent (gravity vent) the peak pressure was 63% above the allowable peak pressure for the structural strength set, and could result in structural failure.

The discharge test results showed that the dynamic co-efficient or RTOP factor for each vent was:
 

VdS calculation software co-efficient review
From the data provided and the results of live IG55 and FM200 discharge tests, we can show that the dynamic co-efficient of a pressure vent is as critical to the calculation as the calculated FVA.

As discussed above, if we just calculate system pressure venting using FVA only, the predicted maximum peak pressure will be false. The performance characteristics of the vent to be used must be included in the calculation as a factor to accurately predict risk peak pressure.

The VdS calculation software already includes this, as the factor used for the vent is variable and the correct factor as detailed above can be imputed so that the correct FVA figure is provided.

The data output from the VdS software will still be a FVA figure but by using the correct dynamic co-efficient sizing, the vent using the output FVA will result in the correct sizing and a more accurate peak pressure being achieved.

If we look at the peak pressure results detailed in the discharge test results and discussed above, the actual sized pressure vent required to keep the risk below 500 Pascals would be:

This shows that a FVA figure can be used but must be calculated using the correct dynamic co-efficient to provide the correct sized pressure vent.

Damage can be caused to the fabric of the building and could compromise the effectiveness of any fire stopping

Conclusion
The sizing a pressure vent for a gaseous fire extinguishing system is currently calculated with very little concern or focus on the performance of the type of pressure vent selected for the system. To date, there is only one software package that includes a factor for performance of the vent.

When this factor is used correctly, the resulting FVA calculated will relate directly to the size of vent to be used and can result in vast size differences.

The sizes can vary from a 200mm x 200mm vent to a 600mm x 600mm sized vent, with the resultant risk peak pressure remaining the same.

Using the dynamic co-efficient when calculating free vent area is critical and must be included in all venting calculations to avoid structural damage or damage to fire stopping and other areas of the risk construction, and affecting the agent hold time.

The dynamic co-efficient figures obtained in the test data are:

If these figures are used for calculating venting and peak pressures when system design is carried out, the resulting stress on building structures will be reduced dramatically.

It is also imperative that the fire protection industry implements a testing protocol for manufacturers to test their vents to. This cannot be left to the venting industry, as too many of them have no or very limited understanding of gaseous firefighting systems and the dynamics required.

Recommendations
Manufacturers of vents should have their vents tested at three pressures using live agent discharge test criteria. In April 2008 AFP Air Technologies produced a live discharge testing protocol which was verified and witnessed by BRE. This provided a dynamic co-efficient for the three vents tested based on an open hole of 300mm x 300mm giving a FVA of 0.09m². This protocol was set out so that it could be refined and used by the industry to test any type of pressure vent offered.

Tests cannot be carried out by using fan pressure in a box or duct, as this will not provide the dynamic co-efficient of the back pressure/risk peak pressure produced by the effect of the vents’ resistance to opening pressure in a live discharge.

The results of this live performance test will be very different for some types of vents. For example, a gravity type vent will have a very different dynamic co-efficient at 250 PA to its dynamic co-efficient at 150 or 500 Pascals, where as a counter-balanced vent will have the same co-efficient from the point it is fully open. The only problem with testing this type of vent with a fan is that it may be fully open at 100pa but when tested in a live gas discharge, the risk peak pressure created from the resistance to opening pressure will be higher.

The test protocol should use three open holes sizes to determine the dynamic co-efficient, and these holes must relate to free vent areas set at 150 Pa, 250 Pa and 500 Pa and provided by the agent manufacturers.

By using these three pressures, a curve will be produced for the dynamic co-efficient as this figure may and will probably be different at each tested pressure. From this data, the risk peak pressure calculation can then be accurately made using the particular dynamic co-efficient for each vent.

As many chemical agents discharge tests are restricted, I would propose that the test data from the inert gas tests, which are worst case, are used for any calculation for chemical clean agents.

Paul A Coxon is principal of AFP Air Technologies LLP

References
1. Experimental Studies of the effects of Blasting on Structures by A.T. Edwards & T.D.Northwood

2. Analysis of Building Collapse under Blast Loads by B.M. Luccioni, R.D. Ambrosini & R.F. Danesi

3. Fragility Assessment of Light-Frame Wood Construction Subject to Wind and Earthquake Hazards by Bruce R. Ellingwood, David V. Rosowsky, Yue Li, and Jun Hee Kim;

4. Pressure Dynamics of Clean Agent Discharges by Mark L. Robin, Eric W. Forssell and Vimal Sharma.

5. AFP Pressure Vent Test Report 2008, Verified by The Building Research Establishment, By P.A. Coxon.