Edit this article

Last edited 15 Apr 2025

Main author

CIAT Website

Moisture, fire safety and emerging trends in living walls

1 How wet is your wall? Exploring fire safety and emerging trends in the use of green walls on tall buildings
2 When is a class B, not a class B?
3 The lack of adequate monitoring
4 Take the living wall challenge
5 Myth busting
6 “We can hook up the living wall irrigation system to the fire alarm”
7 “We have done this on dozens of projects, of similar height and size to yours”
8 “A healthy wall is a safe wall”
9 The problem with failure
10 Going forward: suitability for tall buildings?
11 Related articles on Designing Buildings

[edit] How wet is your wall? Exploring fire safety and emerging trends in the use of green walls on tall buildings

Architectural and design practices, keen to utilise the many benefits of a green wall on their project are faced with a challenge: add a green wall and you add the increased risk of fire spreading across the outer surface of your building. So how does that work with compliance, regulations and guidance documents? What about the additional design considerations for tall buildings? Green walls, also known as living walls, are finding their way onto larger and taller buildings. However, this trend might seem contradictory considering the increased scrutiny on external cladding products.

The issue remains unavoidable: Plants are combustible. The frequent media coverage of wildfires, and their devastating consequences, would reinforce the notion of rapid-fire propagation in plants that is then very challenging to control.

Designers should therefore carefully assess whether the pursuit of the numerous benefits associated with greening is overshadowing the additional inherent risks related to potential fire propagation inherent to living walls. With added significance in the case of tall buildings, a comprehensive understanding of the risks is essential.

So how then are green walls being permitted for use when other façade products that would very likely perform better in reaction to fire testing are prohibited for use over certain heights and in particular locations? An examination of the fire certification and test reports available from the main system suppliers uncovers a narrative that is not as widely recognised as it ought to be.

It reveals that green walls cannot achieve a full reaction to fire classification when tested in accordance with BS EN 13501-01, which would enable designers and review bodies to determine if a product meets the current guidance as shown in Approved Document B, with regards to height and building use.

Yet, surprisingly, some prominent system suppliers are boldly and unashamedly claiming that their system holds a fully tested classification (typically class B), despite this being unattainable. Unfortunately, this claim is being mistakenly accepted as factual.

The Fire Engineering Assessments (FEAs) provided as evidential support by suppliers reveal that certain aspects of the formal testing criteria, such as conditioning and crucial dimensional consistency, are disregarded. Instead, green walls are soaked and completely saturated before undergoing the key laboratory fire testing, which has a significant impact on the results.

This “ad-hoc” testing approach, as described in detail in each FEA, is by itself questionable, and the “wet and test” method to produce a so called “likely” reaction to fire rating is strictly prohibited for other façade products.

In addition to this, the current FEA approach to fire performance for living walls generates supplementary requirements (as well as additional costs) for both the initial design and then throughout the life of any living wall installation. Any installation that is approved for use on tall buildings, based upon the “likely” classification opinion of an FEA document, must also be: (a) Managed, (b) Maintained, and (c) Continually Monitored to remain exactly in the “perfect” condition, mirroring what was tested in the lab. Otherwise, the Fire Engineering Assessment is deemed invalid, and the reaction to fire performance of the installed living wall is unknown.

Despite persistent inquiries and attending supplier CPD sessions, little evidence has been found to show that all the stringent and fundamental criteria outlined in the FEAs is being given the required high priority or being effectively and correctly communicated to designers by the system suppliers.

Furthermore, recently published videos showcasing extended fire testing of living wall systems have unsurprisingly shown that plants burn well. Even for green and healthy vegetation, as the heat from the fire source dries parts of the wall above, the spread can be very rapid. This has significant shortcomings in the existing testing procedures. These videos suggest that the third-party Fire Engineering Assessments, used to permit living wall use on tall buildings, rely on limited data from small scale tests, which do not necessarily reflect real-world installations.

While the importance of maintenance is championed, suppliers are not addressing the key requirement of accurate monitoring of moisture levels within the installed plant modules. This monitoring obligation is being explained as “you could do it”, rather than “you must do it”. The stipulation of constant monitoring of moisture levels is critical to fire spread, but this is not typically being implemented on each and every installation, as is required by the FEA.

[edit] When is a class B, not a class B?

A review of the available systems on the UK & Ireland markets, shows that a typical “likely reaction” result, via an FEA for a fully saturated living wall, would be Class “B” with some variation in the Smoke designation (s1,s2,s3) and the droplets designation (d0, d1, d2) between the different systems. This variation is dependent upon the plants, the material of the plant modules (commonly plastic) and the growing mediums used.

As per table 12.1 (top image far right), this immediately rules out potential use of any living walls on Relevant buildings that require a min class A2-s1,d0. But the suggested ‘Class B’ rating of living walls would potentially allow use of these systems on the external surface of buildings of any height when the potential building use would be offices, Multi Storey Car Parks and some low rise residential. This is where we see living walls on many of the recent installations.

The major problem here is that is that the “Likely class B”, is fully dependent upon the condition of the installation being as what the FEA describes as “Perfect” and then it must remain fully saturated at all times. Any variation of the condition, such as a drop in moisture content below the tested levels, invalidates the FEA.

Examples of fully or partially failed walls, as depicted in the image of the living wall on the next page, are not uncommon due to the inherent challenge in all vertical planting of retaining moisture and preventing local drying out of the growing medium.

In the image (above top right), the portion labelled “A” may potentially meet the ad-hoc test criteria and attain the minimum moisture level for the system; however, without an active method of accurately monitoring percentage moisture content (%MC), how could you tell?

Furthermore, a fire emanating from the window below could rapidly decrease the moisture content and impact the reaction to fire performance. This would not typically be the case with other cladding products subject to the full BS EN 13501-1 criteria.

A further key question is: Are the moisture levels across the wall on this particular project example and any others accurately known? If not, why not?

The opinion given in a fire engineering assessment stands or falls upon the minimum moisture content being met. If a section of this living wall was to be tested, it would very likely not achieve a “Class B”. Is this particular wall, and any others with similar issues, therefore compliant with regulations in its current state?

The conditioning problem is not always visually obvious which is another reason why the saturation method of testing is problematic. An apparently healthy-looking wall with a lower moisture content will enable significantly quicker fire spread as it is the moisture within the system that provides protection from fire for both the foliage and the commonly used plastic plant support modules.

Side by side testing carried out by Prof Wojciech Wgrzyski, Jakub Bielawski and the team at ITB shows that there is a significant increase in fire spread across living walls with a marginally lower moisture content or for a wall rapidly dried and fuelled by the wind (see images above right). This confirms that a detailed and specific green wall testing regime is very much needed.

The videos indicate that the flames quickly spread across the face for each of the samples tested, with the differences being limited to time of flame spread. These are much more realistic conditions for green walls when used externally and at higher levels. Within minutes of the tests starting, there was significant fire spread as shown in the images (above right).

This raises additional doubts about the practice of omitting crucial aspects of BS EN 13501-1 testing and saturating a wall to attain a so called “Class B” rating which has clearly opened the door (and the market) for use of green walls on tall buildings. The Luton MSCP fire has also raised doubts about the common practice of green wall installations on car parks via an apparent Class B. Compared to other façade products achieving a full class B rating, living walls have been shown to be significantly inferior in preventing fire spread. It is evident: green walls have the potential to ignite and burn effectively.

As the reaction to fire properties of the living wall system can rapidly change as demonstrated in the larger scale tests of both wet and dry walls, the stated “likely class B” designation for these systems would appear to be dubious.

[edit] The lack of adequate monitoring

Construction products given a fire classification rating are expected to perform as tested in a fire even if fully dry. As living walls will perform differently dependent upon the moisture content, the FEA will stipulate that they must be “constantly monitored” and irrigated as required to prevent them from falling below the minimum permitted level. The minimum stated levels for saturated ad-hoc tested walls varies depending on the growing medium used, with some systems being as low as 40% minimum moisture content (soil) and some being closer to 70% of volume (mineral wool).

The lab method of doing this (measuring by weight of modules before and after saturation) is not practically possible for almost all project installations, so therefore some other method is required to confirm the actual percentage level. It is worth noting that test labs set the minimum levels, but no guidance is given as to how this should be achieved.

The current monitoring method that the main system suppliers suggest is computerised management of flow rates from the irrigation system, and this is now typically standardised across most of the available marketed systems. This method does not confirm the exact moisture level of the substrate itself or give accurate readings of moisture content percentage levels (%MC). Flow rates can be finely tuned, with published daily data readings being produced for the volume of water being pumped to the wall that can be remotely adjusted if required. While this form of monitoring is essential in plant health, it falls short in meeting the requirements of the Fire Engineering Assessments that set a minimum level.

This is because all current monitoring happens primarily ‘bit side’ as per location (B) on the below diagram (left), which represents a typical installation. Flow rate monitoring happens on the irrigation feed pipes for the system, usually located in the plant room or on the supply lines.

While a potential issue with a leaking, damaged pipe or blockage may be flagged up by a pressure drop indication when using this method, there is not an accurate method being applied to most installations to constantly monitor the wall itself, within the plant modules. This should happen at location (A) to confirm an actual moisture figure, (%MC) which is a clearly determined stipulation of the validity of the supporting FEA document.

As the fire engineering assessments set out a minimum level in percentage terms that the installed wall must achieve (as per testing), again logic would suggest that to understand if this minimum level is being met for each project, the actual moisture level (%MC) in the wall must therefore be known. It is typically not.

Advanced moisture sensor technology is commonly employed in commercial agriculture and vertical farming to manage plants. However, through research and conversations with suppliers, this technology is generally not being implemented in most living wall installations. When the moisture percentage is not monitored accurately, the probable fire reaction classification for the wall also remains uncertain and potentially non-compliant.

A straightforward comparison to highlight the shortfalls in flow rate monitoring would be the process of refuelling a typical car. While each fuel pump precisely measures the amount of fuel in litres passing through the nozzle into the car’s tank, this measurement is entirely independent of consumption, impacted by the journeys you may undertake.

On a lengthy trip, your total fuel consumption will be much higher compared to a short one. If you attempted to maintain the fuel quantity above 50% (a half tank) in your car by solely pumping in a fixed amount of fuel each morning, you would encounter difficulties as this approach doesn’t consider consumption. Fortunately, cars are equipped with a fuel sensor in the tank, providing real-time and accurate feedback.

Similarly, merely knowing the volume of litres passing through a living walls irrigation system doesn’t account for local daily environmental conditions that can reduce moisture levels within the installation. While you could try to make an estimation, without a precise measurement method, you simply cannot have certainty. FEAs require that a certain minimum level is always met.

For designers, it is important to understand that the reaction to fire of a living wall system can be impacted by local environmental conditions such as dry weather and wind. As the moisture levels are reduced, so does any protection that this gives against fire spread. This may not always be obvious to spot.

[edit] Take the living wall challenge

Just how wet is your wall? If you have been involved in a project with a living wall that has been approved for use on a tall building via this non-standard ad-hoc testing (as detailed within each system’s FEA), then the likelihood is that this is fully dependent upon the moisture content remaining above a stated minimum level. If so, ask what the moisture content (% level) of the living wall is today? This should be known by whoever is managing the wall and should be relatively easy to find as it is the critical element of the Fire Engineering Assessment.

Is this level above the minimal levels in the systems FEA? These documents should also be readily available. If this information is not readily available or known for your project, the FEA for the system is likely invalid and the reaction to fire classification for the installation is therefore not known.

It is worth noting that Approved Document B (ADB) refers to an additional document for best practice guidance which is the “Fire Performance of Green Roofs and Walls, published by the Department for Communities and Local Government” document (published Aug 2013.)

A careful read of this document shows that it contains some outdated information and diagrams which have been superseded in recent updates to ADB. As is now widely being highlighted by fire safety specialists, this documents overall relevance today is questionable, and the contradictions contained within it are unhelpful to those tasked with reviewing green walls today.

Rather than providing additional and much needed clarity, this document adds to the confusion. It notes that the green walls on the market at the time of writing and that were tested before this publication, all failed the relevant testing, and that further guidance is therefore needed.

[edit] Myth busting

There have been various ideas suggested to address the logical fire concerns that come when putting known combustible materials onto the outside of buildings. Some of these appear to have stuck and have been incorporated within recently installed green walls. But do they satisfy the current regulations and guidance or are they just a good idea?

[edit] “We can hook up the living wall irrigation system to the fire alarm”

The idea here is that in the event of a fire, the building’s fire alarm activating would send a signal to the irrigation system. This would then irrigate or “drench” the wall which would then reduce its likelihood to catch fire.

The obvious question here is: where is the testing to prove that this works? Any building, or part of a building, that is reliant on a sprinkler or other fire suppression system is required to be designed by qualified and experienced specialists and be certified and commissioned. Green wall irrigation systems are pressurised drip systems to evenly distribute water and nutrients to the plants in a controlled manner. They are not proper fire suppression systems. While there is a likely benefit, this is not a requirement of the regulations for external facades, nor is it an alternative pathway to allow products that fall short of the test standards to be used.

Also, if the wall meets the required reaction to fire performance in the first place, why would this even be suggested by suppliers? The suggestion to do this in fact further highlights that the inherent risk with living walls becoming a source for fire spread, is a major problem.

[edit] “We have done this on dozens of projects, of similar height and size to yours”

While this once held much weight, the industry must now be fully aware that just because something was done (or permitted) on a previous project, this is no guarantee that those involved in the approval process got everything right, especially with regards to fire.

It is imperative for all stakeholders in the design process to thoroughly scrutinise and question both living walls and the associated documents submitted for approval, ensuring a comprehensive understanding of the potential risks. Suppliers should refrain from overstating the “likely” classification achieved through the Fire Engineering Assessment and ad-hoc testing process in both discussions and publications.

Designers must fully understand the risks and FEA stipulations to be able to make informed design decisions. They must comprehend the implications of deviating from standard fire testing procedures of products and the complete reliance of moisture within the system to limit fire spread.

[edit] “A healthy wall is a safe wall”

This simplified claim would suggest that maintenance contracts are therefore the silver bullet for fire safety. While regular maintenance is essential for plant health,

it alone falls short of showing regulatory compliance. Current maintenance contracts include live flow rate monitoring of the irrigation system (during work hours) and consists of a monthly or a bi-monthly visit to do a full visual inspection of the installation and irrigation equipment. This tends to be quite thorough and includes the replacement of any unhealthy or dead plants as and where necessary. It helps, but still falls short of carrying´out the FEA stipulation of “constant monitoring of the moisture levels” to confirm the likely reaction to fire performance of the system.

Variations in weather and other factors can influence the health of the wall, even over a short timescale, which is why maintenance is needed in the first place as some parts of the wall may not thrive. As demonstrated by full scale testing, even healthy-looking walls, when subject to a realistic fire load, will quickly dry out, enabling fire to spread rapidly over the surface. The more moisture, the slower the spread as the plants will take longer to dry but overall, spread is only slowed rather than “resisting fire spread” which is the overarching intention of B4.

This simplified assertion being put forth by suppliers that “a healthy living wall is a safe living wall” currently lacks substantial evidential support. This notion parallels the “it does not burn, it just chars” claim made in relation to insulation products, which is now proven to be misleading. For designers tasked with navigating the complexities of these systems and fire safety, caution must be taken with such claims.

[edit] The problem with failure

Green walls can fail and inevitably we will see increased examples of this as more are installed. So where does that leave us?

Living wall system suppliers are rightly passionate about their installations looking good and will endeavour to secure the maintenance contract. Unfortunately, they are also in a position where they cannot guarantee that their maintenance contract will be renewed or that a client will not “turn off the tap” for some unknown reason. This creates a unique and bizarre situation with regards to regulatory guidance for living walls.

The reality is illustrated through these contrasting images below showing a living wall in its first year and, the same wall recently.

Once the jewel of this mixed-use and multi award winning community project, for whatever unknown reasons within the last twelve months, the irrigation system appears to be switched off and no maintenance has taken place. While it is undeniably disheartening to witness these outcomes, this project, along with some other prominent and more widely publicised cases, do serve as a reminder that although failed living walls may be relatively uncommon, there remains the potential for them to deteriorate and dry out: factors beyond the control of system suppliers.

At times, either the entirety of a living wall or a specific localised sections may not thrive, and the result will be a highly flammable outer surface. This situation presents a challenging dilemma concerning regulatory guidance, necessitating further clarification to for allowing parts of a building to deviate from the requirements set out in Approved Document B.

[edit] Going forward: suitability for tall buildings?

Urban greening is important, and living walls clearly offer an excellent means of achieving this goal. However, the current pathway permitting their installation at any height on certain buildings (via table 12.1 see top image right hand side), using the wet-it-and-test-it method to achieve a purported “likely class B” designation, requires further scrutiny.

Additionally, designers need clear guidance on how ad-hoc “likely classifications” should be managed. Presently, some consider them to be fully equivalent to a BS EN 13501-1 tested product, a perspective that seems to have inherent flaws, especially for living wall systems, given the potential for a rapid decline of fire performance linked to change of condition, as evidenced in the larger scale testing.

Furthermore, it is crucial to recognise that the current design review scenario is inherently flawed, where both designers and review bodies are heavily reliant on the claims being made in suppliers’ own documentation for these so-called Class B systems, but do not then have clear regulatory guidance on how this information should be interpreted.

A review of living wall installations on tall buildings to confirm that all stipulations within their system FEA are being met is a must.

For the time being, it may be prudent to limit the design and installation of Living Walls to areas where the associated risks are less. This approach could prioritise projects such as low-rise buildings, creating distinct breaks between planted areas and avoiding use on taller buildings altogether, despite the potential unpopularity of this.

This cautious approach should remain until further testing, clear and comprehensive guidance, along with specific regulations, have been firmly established for living walls.

Numerous opportunities do exist at lower levels on most construction projects for incorporating urban greening, and there are numerous recent successful examples to draw inspiration from.

Based upon recent larger scale testing, it is evident that urgent changes are necessary. Those involved in the approval process must be adequately educated about the associated risks and the limitations with the current lab tests.

Regarding design, plant support systems should be reviewed to remove any combustible materials or components from both the support framing and the planting systems, replacing them with non-combustible alternatives. This will reduce reliance of moisture within the system for protection. Some emerging systems on the market that prioritise fire-safe designs for the entire system are currently doing this.

With the necessary talent, technology, and expertise within the industry, there is a strong potential to develop a well-defined large scale testing process and a set of much needed standards for green walls. This testing should also identify and provide design guidance for external breaks and barriers within the planting.