Barriers to the Effective Implementation of Performance-Based Design

Europe_Digital_Magazine.png

View the PDF here

By Greg Baker

This article is the second in a series of articles that summarise a presentation entitled “The Contribution of the RTM to the Global Advancement of Performance-Based Design,” given by the author at the SFPE 2020 Performance-Based Conference & Expo, which was staged in Auckland, New Zealand, from 11 to 13 March 2020.

 

Introduction

While the conference presentation as a whole focused on the broader role of SFPE’s Standing Committee for Research, Tools and Methods (the so-called RTM) in advancing the implementation of performance-based design (PBD) internationally, this second summary article from the presentation specifically deals with the author’s views on some of the barriers to PBD being implemented effectively.

The context for this second article is a new SFPE committee that was formed in late 2019 to develop a new SFPE standard on PBD. One objective of this new standard-making committee is to reduce the barriers to the effective implementation of PBD by producing an ANSI-accredited SFPE standard on PBD that can be widely used and relied upon around the world in numerous regulatory jurisdictions.

 

For clarity, it should be noted that the views expressed in this article (and the conference presentation) are those of the author alone and do not represent the formal view of the above-mentioned committee or, for that matter, SFPE as a whole.

Defining Performance-Based Design

The first article in this series, “Defining Performance-Based Design,” was published in the Q1 2020 Issue 17 edition of SFPE Europe Magazine.[1] The article provided the author’s definition for PBD and the comparative status of different exemplar New Zealand Building Code compliance documents in relation to the definition.

To recap, the stated definition for PBD in the first article was:

PBD methods are alternative, risk-informed methodologies used to demonstrate compliance with fire safety objectives


In the first article, the author noted that the definition was one possible way of defining PBD in the context of performance-based fire safety engineering.

Under this definition, for a design method to qualify as a PBD method, the design process must use methods that are alternative to the deemed-to-comply methods prescribed by the regulator (such as deemed-to-satisfy provisions or verification methods).

In the context of the stated definition, a risk-informed design method must utilise probabilistic analysis methods (also termed quantitative risk analysis or QRA).

As also noted in the first article, the author acknowledges that the definition for PBD is consciously narrow (for example, qualitative risk analysis methods are excluded) and specific, whereas different stakeholders will often have broader goals and objectives.

Barriers to the Effective Implementation of PBD

There are a number of barriers to the effective implementation of PBD, seven of which are described in the subsequent paragraphs.

Barrier 1: Legal/Regulatory. The most basic barrier to the effective implementation of PBD is the need to have either a performance-based building code or provisions in a non-PBD code that permit PBD as a design option in the first place. If the building regulatory framework in a jurisdiction does not even permit PBD, this is a significant and fundamental barrier to PBD being effectively implemented in the jurisdiction.

Barrier 2: Definitional Clarity. A lack of understanding of what is actually meant by PBD is also an important barrier to PBD being effectively implemented. This is particularly relevant where the design method in question is considered to be PBD, when in fact, it actually is not PBD.

Barrier 3: Sector Capability. The expertise and competence of various practitioners in the design/approvals chain is another significant barrier. If, for example, the building official from the authority having jurisdiction (AHJ) does not have sufficient expertise/competence to review and approve a performance-based design, but goes ahead and issues approval, unsafe design may result. This results in possibly compromising life safety as well as undermining the credibility of PBD as a whole. In this scenario, heavy reliance is placed on the designer to “get it right.” If the designer also lacks the requisite expertise/competence, this has significant potential to undermine the practice of PBD.

Barrier 4: Quantification. A recurring theme in countries where performance-based building codes exist, even in some cases for decades, is that a lack of quantification generally in building code clauses (i.e., the clauses are qualitative or semi-quantitative in nature) hampers the practice of PBD in that jurisdiction. A typical response to this situation is to use prescriptive, deemed-to-satisfy provisions as the yardstick for PBD — a process sometimes termed “equivalency.” The problem with this work-around is that the prescriptive provisions are often not quantified in a manner suitable for PBD, e.g., the prescriptive provisions can be very specific, which does not lend itself well to a comparative PBD approach.

Barrier 5: Probabilistic Acceptance Criteria. Accepting the definition for PBD given in the introduction to this article, another aspect that hinders the practice of PBD is acceptance criteria that are not expressed in a probabilistic manner. The definition stated previously is based upon PBD needing to incorporate so-called “risk-informed” design methods. To facilitate PBD, therefore, relevant acceptance criteria need to be expressed in a way that unequivocally quantifies the probabilistic thresholds of acceptability.

Barrier 6: Accepted Tools and Methods. To be able to implement probabilistic analysis as part of PBD, design tools and methods are required. The majority of design tools and methods are deterministic (i.e., non-probabilistic) in nature, so a lack of probabilistic tools and methods is not a trivial operational barrier to PBD being implemented more widely.

Barrier 7: Societal Impact Barrier. A very important political and/or societal barrier to the implementation of PBD is the implicit acknowledgement that injuries and/or fatalities will possibly, albeit rarely, occur. This is, of course, not to say that PBD is any different to prescriptive design, it is just that the issue is overtly transparent in PBD where probabilistic methods are utilised.

Examples of Barriers

In relation to Barrier 4 noted in Section 3, an example of a barrier is Clause C2 – Prevention of Fire Occurring in the New Zealand Building Code (NZBC), where the Functional Requirement Clause C2.1[2] states:

C2.1 Fixed appliances using controlled combustion and other fixed equipment must be designed, constructed, and installed in buildings in a way that reduces the likelihood of illness or injury due to fire occurring.

The wording of this code clause in the NZBC can be described as being “partially risk-informed” in that the probabilistic term “likelihood” is used. However, the specific nature of the barrier in relation to quantification is that no actual quantification of the likelihood is provided by the phrase “reduces the likelihood.” The problem for both the designer and the AHJ official is a need to subjectively interpret the quantum of the reduction required, e.g., in this case, a design with either a 1% reduction or a 90% reduction would still meet the criterion, but obviously with a vastly different level of fire safety being delivered.

Again, in relation to Barrier 4, another example from the NZBC is Code Clause C4 – Movement to Place of Safety, where Functional Requirement Clause C4.2[2] states:

C4.2 Buildings must be provided with means of escape to ensure that there is a low probability of occupants of those buildings being unreasonably delayed or impeded from moving to a place of safety and that those occupants will not suffer injury or illness as a result.

As with the previous example, this NZBC clause can again be described as “partially risk-informed” with the use of the term “low probability.” In the context of Barrier 4 from the previous Section 3, the terminology “low probability” is non-quantified and hence subjective. Once again, a designer or code official must make a subjective decision of what constitutes “low probability” — for example, does this mean a 0.1 chance of success (i.e., 1-in-10 or 10%), or a 0.01 (1-in-100 or 1%) chance of success, or a 1 x 10-6 (1-in-1-million or 0.0001%) chance of success, etc.?

The barrier in both these cases of code clauses extracted from the NZBC is that the “partially risk-informed” terminology used in both code clauses is not quantified sufficiently to prevent the designer or AHJ official needing to make a subjective judgement on what either “reduces the likelihood” or “low probability” means. Making a subjective judgement of this nature transfers risk to the person making that decision.

The two examples provided so far of barriers in the NZBC are both examples of Functional Requirement clauses. The set of Protection from Fire Functional Requirement clauses in NZBC are generally qualitative, while the associated Performance clauses provide a greater level of quantification.

For example, one of the Performance clauses associated with Functional Requirement Clause C4.2 states in Performance Clause C4.3 (a)[2]:

C4.3 The evacuation time must allow occupants of a building to move to a place of safety in the event of a fire so that occupants are not exposed to any of the following:

  1. A fractional effective dose of carbon monoxide greater than 0.3:
  2. …
  3. …

In Performance Clause C4.3, tenability criteria are explicitly quantified. Although quantification is provided (Barrier 4 identified previously), the clause (which essentially consists of acceptance/performance criteria) is not risk-informed (Barrier 5), so hence does not comply with the stated definition for PBD.

To counter this view, one might say, for example, that “the regulator has determined the societally-acceptable level of risk is a fractional effective dose (FED) of 0.3 and thus Code Clause C4.3 is risk-informed.” However, this does not fully meet the definitional “risk-informed” threshold for PBD, with the element still missing being an explicit probabilistic component to the quantification.

Key Needs to Reduce Barriers

Having identified barriers to the effective implementation of PBD and provided specific examples from one jurisdiction of how such barriers are manifest in building regulation, the obvious question to pose is “what are the key needs for the international fire engineering community to reduce these barriers?”

To advance the practice of PBD globally, the seven identified barriers can be distilled down to three key needs that, if addressed appropriately, will reduce barriers to the implementation of PBD.

Key Need 1: International Acceptance. The first key need is for comprehensive design methodologies to be developed by authoritative organisations that have the international mandate to do so. This is  an important first step to such design methodologies having sufficient credibility to be the precursor for widespread international acceptance occurring.

Key Need 2: Quantified Design Criteria. On the basis of the definition presented for PBD, the second key need is for performance and acceptance criteria to be quantified with the combination of specific target values and an acceptable probability of non-exceedance for each target value.

Key Need 3: Tools. The third key need is for suitable computer tools to be developed, validated, introduced, and supported so that practitioners are able to consistently implement probabilistic design criteria in their design analyses.

A planned future article will provide examples of how SFPE is, and will continue to be, reducing the barriers to the effective implementation of PBD.

References

Greg Baker is with Fire Research Group Limited, New Zealand.