Advertise on HVAC&R Online
Advertise on HVAC&R Online
Advertise on HVAC&R Online
Advertise on HVAC&R Online

Ventilators: when is a zone not a reservoir

By Ron Burns

We take a look at the requirements of a smoke zone versus that of a smoke reservoir.

The roof-mounted smoke ventilator received a bit of scrutiny last month. We are still performing a range of tests to see how much we can load the fixings. It will be good to finally put this highly debated topic to rest.

As promised, the testing and results will follow as soon as we have our test slot available. I want to keep the discussion on roof ventilators open. Roof ventilators are topical in the smoke sector of the industry with the summer rains and high winds beating down on South Africa at present.

EN 12101 Part 5 is quite specific about reservoir sizes: Where the fire is directly below the smoke reservoir, the maximum area of any one reservoir should be 2 000m2 if natural smoke ventilators are fitted.
[Technical Report prCEN/TR 12101-5]

The Code goes on to mention that the longest boundary of the smoke reservoir cannot exceed 60m. There are warnings regarding buoyancy — best ignored it seems, judging by the tender documentation passing my desk. I think it is a genetic deficiency: I see many tenders with extended smoke zones. I think the approach is similar to the South African driving protocol: stop signs are treated as yield signs; a green traffic light … best you stop and check, as the person approaching the red traffic light on the opposite end (as the standard rule of thumb) probably has no intention of yielding, let alone stopping.

The all-too-often adopted stance of using the doors as the replacement air path, falls short of compliance should the fire occur when the building is not trading and/or occupied.

Therefore, if both boundaries are 60m, then 60m × 60m = 3 600m2, so “The code must be wrong” approach will suffice. For those pondering the economics of the associated mathematics ... Financially speaking, a single smoke zone with ventilators, in today’s terms, is approximately R160 000 (excluding VAT). Keeping with the standard value engineering principle, replacement air shall be ignored, and the necessary savings passed on. My concern is the risk and that this never gets passed on. The excitement at passing on a saving and assuming the risk of extend reservoir boundaries never ceases to amaze me.

3.1.34  Replacement air
Clean air entering the building, below the smoke layer, to replace smoke gases being removed by the smoke and heat exhaust ventilation system.
[Technical Report prCEN/TR 12101-5]

Picturing a 120m × 60m = 3 600m2 building, based on the logic above, this would result in two smoke zones and a smoke baffle — all for little more than double the R160 000 budget for the above-mentioned building. The building would appear functional at first glance, with replacement air supplied via the adjacent smoke reservoir — a better chance of working than the 60m × 60m building. There now is replacement air and I would think (as an educated guess) that the dual smoke zone warehouse (120m × 60m) would perform better in the event of a fire than the single zone option (60m × 60m).

The concern I have with the installation is that the building should be segmented into four smoke zones. Besides a saving to the client (I assume a little over R360 000), the building is at risk of complete loss of product in the event of a fire, and the lack of buoyancy would result in excessive smoke logging. Unfortunately, I cannot substantiate the result of the extended smoke zones; my thermodynamics are a little rusty. I think it would be far better to assume that the Code is correct, and that the smoke would lose its buoyancy. This would result in smoke cooling and sinking due to the density differential versus replacement air. This low-level smoke would be drawn back into the seat of the fire. The result is excessive smoke generation. This, coupled with the lack of buoyancy would render the smoke ventilators’ aerodynamic area too small to ventilate the volume of regenerated smoke.

The problem is simple to solve. Invest in the additional smoke barriers and ventilators, and provide the client with a system that, although a little costly, will work. As a client, I would rather spend R720 000 on a system that is functional than R320 000 on a system that will fail.

Trying to solve the ventilator fixing detail, I stumbled across this comment in part 4 of the EN 12101 code:

4.1.1    Smoke layer exhausted by natural vents
The needed area of vents is measured by the free area for each zone. For efficiency of the Smoke Heat Exhaust Ventilation Systems (SHEVS), the maximum area of a smoke zone shall not exceed 1 600m2.

Air inlets have to be provided to allow for the replacement of smoke by fresh air:

  • With opening in roofs
    The efficiency of the vents is measured by their aerodynamic free area Aw m2.
    There shall be at least one ventilator for 200m2 of floor area.
  • With opening in side-wall
    The efficiency of the vents is measured by their aerodynamic free area Aw m2.
    Wall-mounted ventilators have to be installed on at least two different sides of the building and be monitored by wind direction control system.

[Technical Report prCEN/TR 12101-4]

I have always advocated against the use of vertically mounted ventilators, due to the functionality problems with operating and monitoring the wind direction. But we can leave that topic alone for this discussion.

Another comment in the above extract (which is often overlooked), is the ventilators per 200m2 requirement. With the focus on economics, there is a huge drive to provide the ‘most cost-effective solution’. This is all too often replaced with the term ‘the cheapest solution’. There is a difference between a mathematically compliant system, which can easily be the cheapest solution, versus the most cost-effective solution. As an example, if it is possible to achieve the calculated aerodynamic area by supplying eight ventilators in a 2 000m2 smoke reservoir, then eight ventilators are supplied. I cannot recall the ‘one ventilator per 200m2’ rule being applied. The client now has ‘the cheapest solution’, which is challenging compliancy, versus the ‘most cost-effective solution’, which provides a functional system complete with effective smoke protection.

I read the extract again and noticed the difference in the terminology. Part 5 refers to a smoke reservoir and Part 4 refers to a smoke zone. I think we should look into the definitions and get some clarity on the terminology.

A smoke reservoir as defined in Part 5 states:

3.1.43  Smoke reservoir
Region within a building limited or boarded by smoke barriers or structural elements in order to retain a thermally buoyant smoke layer in the event of a fire.

[Technical Report prCEN/TR 12101-5]

The definition in Part 4 is a little different:

3.28     Smoke reservoir
Volume within a construction works limited or bordered by the ceiling and smoke barriers of structural elements so as to retain a thermally buoyant layer in the event of a fire.

[Technical Report prCEN/TR 12101-4]

Roof mountedFigure 1: Roof-mounted natural smoke and heat exhaust ventilation in a single-storey building with louvres-type ventilator.

Figure 1 offers no clarity relating to the ‘smoke zone’ area. In Figure 1, it is apparent that the building contains three compartmentalised smoke zones; however, the floor level where the fire takes place is not compartmentalised, divided and/or segmented in any way.

Before we dig a little deeper, we need to pause here and reflect on the commonality of the term buoyancy. Perhaps my ranting about extending reservoir area carries merit. Buoyancy is the engine required to drive the smoke from the building. Failure to maintain the buoyancy, severely and negatively impacts on the ability to evacuate the smoke from the reservoir.

3.29     Smoke zone
A smoke zone is a room or a division of a room of a construction works for the extraction of smoke and hot gases. Each zone is served by a SHEVS (or subsystem of a SHEVS), which is initiated by a signal from a single device or group of initiation devices associated with the zone. A zone contains at least one smoke reservoir.

[Technical Report prCEN/TR 12101-4]

At this point, I am none the wiser as to the relationship between the smoke reservoir size and the smoke zone size. The question I am wrestling with is the relationship between a 2 000m2 building and the division thereof into smoke zones. Looking at the reservoir requirements as a stand-alone requirement, I conclude that a single smoke reservoir is required. However, looking at the size limitations of the smoke zone, I contemplate the division of the space into two smoke reservoirs. The advantage of the division into two smoke reservoirs is the solution to low-level replacement air. In Figure 1, it is apparent that the replacement air is controlled using operable louvres. A code requirement, which we can explore another day.

I invite readers to share their views. By the time this goes to print, a forum will be available for comment on the website.

I pose this question: would the installation of two smoke zones and a smoke barrier result in a lower overall cost to the project? I am certain that the complexity of a wind station, monitoring, and driving the replacement air louvres open and closed in terms of the replacement air wind speed direction and velocity regulation would result in the additional smoke zone and smoke barrier being a more economical installation. Another point worth considering is the roll of the velocity of the replacement air. In a situation where the wind is being blown onto the face replacement air louvres, the control requirements to limit the velocity of the replacement air to below 5m/s will be difficult to maintain.

I have no conclusion to draw between the requirements of a smoke zone versus a smoke reservoir. I definitely think this is a discussion point that raises several attention-worthy items: the control of replacement air velocity into the building; one ventilator per 200m2; and the often-ignored replacement-air path in single-zone buildings. The all-too-often adopted stance of using the doors as the replacement air path, falls short of compliance should a fire occur when the building is not trading and/or occupied.


I invite readers to share their views. By the time this goes to print, a forum will be available for comment on the website. Please comment and contribute — the time to find solutions and effectively prevent unnecessary loss in a fire condition in South Africa’s buildings is now.

Click below to read the February 2018 issue of RACA Journal



06 Jun 2018
08:00AM -
FRIGAIR Expo 2018 - Day1
07 Jun 2018
08:00AM -
FRIGAIR Expo 2018 - Day2
08 Jun 2018
08:00AM -
FRIGAIR Expo 2018 - Day3

logo header



IMD logo White

Interact Media Defined (IMD), is one of South Africa’s leading multi-media magazine publishers READ MORE



CLA Online logo white MM


PLUMBING Online logo white 200


building LOGO white

Talk to us

JHB T : +27 (0) 11 579 4940
CPT T : 0861 727 663
E :

13A Riley Road, Bedfordview,
South Africa 2007

© Interact Media Defined