Resetting the Asset Management Paradigm

1.       Introduction

The following paper was prepared by:

Steve Kent, MM, Ec.D, CAMP

Milos Posavljak, PhD Cndt., P.Eng

It provides a concise summary of detailed asset management study over the last five years as it applies to provision of public infrastructure. The resulting findings and their implications on the industry’s practice are presented. 

2.        Purpose

The purpose of the paper is to provide: 

·        Guidance on proper academic application of asset management principles to provision of public infrastructure;

·        Guidance on proper practical application of asset management principles to provision of public infrastructure.

3.        Academic Reset

In order to successfully study public infrastructure asset management, a full portfolio analysis is necessary rather than individual or groups of only similar asset classes (e.g., only linear). This is because individual groups of infrastructure “become” an asset to the modern community only when they “work together.” For example, having a community with roads but no water and sanitary services, town hall, parks, fires service, recreational facilities, fleet, etc., is not feasible or sustainable.

Due to industrialized world’s necessity for division of labour the same phenomena has been reflected in academia. For example, engineering programs at educational institutions typically include specializations in geotechnical, hydrological, transportation, structural, mechanical, computer, and software engineering. The challenge arises from municipalities being tasked to manage portfolios of assets that include engineering design from all of the aforementioned.    

When studying the application of asset management to provision of public infrastructure the portfolio should be divided into only two overarching categories: non-repairable and repairable assets, rather than the overarching division being the numerous asset classes of roads, sidewalks, trails, bridges, culverts, sanitary pipes, water pipes, storm pipes, storm water management ponds, creeks, parks, parking, facilities, fleet, fire services, and information technology. This is important, as now engineering risk and reliability principles can be applied to forecast future spending and infrastructure performance.

The first step is realizing that out of the 16 asset classes listed above only a few fall under the category of non-repairable systems. These are fleet, information technology, and mechanical systems (e.g., HVAC, equipment) under the asset class of facilities. These typically make up less than 5% of municipal portfolio’s value, while the other 95% is dominated by civil and environmental engineered systems (hereafter referred to as civil engineered).

The main goal of asset management is to determine the optimal timing and cost of future infrastructure improvements. In order to do so forecasting infrastructure performance and corresponding expenditure is essential. It should be noted that current industry emphasis is on data collection (e.g., physical condition of assets) to the point where forecasting and its implications appear to have been overlooked. Specifically, data collection and forecasting methods are equally important for proper asset management; however, the former dominates asset management literature/discussion over the later, thereby upsetting their equal importance to the inexperienced eye. 

Provided the vast quantity of public infrastructure, data collection typically requires significant staff time and/or resources. If improper forecasting methods are used the data collection effort is compromised. It should be noted that continuous annual data collection efforts are typically not financially sustainable for municipalities, hence the further need and importance of proper forecasting.

According to latest research at the University of Waterloo, improper forecasting overestimates needs (i.e., poor performing portion of asset portfolio) by approximately 200% on a full portfolio basis. This can lead to unnecessary tax and user fees increases, thereby negatively impacting the community’s socio-economic wellness. 

Improper forecasting involves applying non-repairable system logics to repairable systems. The non-repairable category consists mostly of mechanical engineered assets (e.g., fleet, equipment, HVAC). Their designers and manufacturers have the ability to do full-scale testing to determine their Estimated Service Life (ESL) before selling them to consumers. At the end of the ESL, these assets are typically replaced at full cost plus inflation, if it is like for like replacement, hence the “non-repairable system” term. This is not the case with civil engineered assets. For example, although pipes are technically replaced, the pipe itself is not an asset for the community unless it is connected to all of the other pipes, thereby making the pipe network (e.g. sanitary network) which serves the community as one asset. Therefore, the proper interpretation of a pipe replacement is the repair or treatment of a pipe network, hence the “repairable system” term. Also, for repairable systems ESLs are representative of meant time to treatment (repair) rather than replacement.           

While the ESL for non-repairable systems is based upon actual performance data (e.g., full scale testing), the ESLs currently used in the industry for repairable systems appear not to be. Rather they are arbitrary numbers created out of necessity to satisfy the Public Sector Accounting Board’s regulation 1350 (PSAB 1350) for municipal Tangible Capital Assets (TCA). While TCA schedules of depreciation have been used for asset betterment (expenditure) and replacement (expenditure) planning for decades in the private sector, this is not the case for the public infrastructure sector. Since its introduction in 2008, it has become evident over the years, and is clear today, that it cannot be used for civil engineered assets making up typically 95% of a municipal portfolio. The key reasons include:

-         Practical relationship between the corporation and its assets

-         In service environment uniqueness

-         Scientific limitations of forecasting

In the private sector, assets typically help the corporation achieve its primary goal. For example, a factory building allows a corporation to manufacture products to be sold on the market.  This is not the case for public infrastructure; in fact, the provision of infrastructure itself is the primary goal for a municipal organization (the absolute majority of capital budgets are spend on public infrastructure). Thereby depreciating them through a prescribed schedule is the equivalent of a manufacturing company carrying on its books and depreciating its products that have already been sold on the market and reached the end consumer. The local community, province, and country are the end consumers of public infrastructure. In order to properly forecast infrastructure needs an adjustment to current TCA methodology is necessary.

Specifically, ESL cannot be prescribed but rather must be derived for each community. Unlike mechanical assets, each civil engineered asset is subject to a unique environment under which it fulfills its purpose. For example, a service truck is designed and manufactured to operate for a certain amount of cycles (i.e., mileage). Its design parameters are not necessary constrained by the uniqueness of the communities it will serve. Hence, when tested at full-scale in manufacturer’s laboratories, the resulting ESLs typically correspond to those experienced in service. Civil infrastructure cannot be tested a full-scale before constructing in the environment. As such, its ESLs and average treatment (repair) costs must be derived from the environment it exists in. This includes analysis of past and current corporate information (i.e. engineering, financial, administrative). The resulting ESLs and average treatment (repair) costs are tailored and unique to the unique community they serve. Alternatively, current TCA schedules paint all municipalities and their assets with the same brush. This lacks scientific backing and therefore requires appropriate adjustments.

Finally, with respect to scientific limitation of forecasting, it is not possible to reliably forecast on a singular basis. For example, car insurance companies cannot reliably forecast at an individual customer level if and when an accident is to occur, what the claim will be, and the magnitude of the claim amount. The same is true for other private sector industries, such as banking and consumer packaged goods.  This is not a limitation of the industries themselves but rather a naturally occurring limitation of probability theory. Reliable forecasting can only occur at an aggregate level. In the case of car insurance, past and current information on all clients is analyzed and then aggregate forecasts – that can further be broken down into specific smaller but still aggregate groups - are used to make today’s decision for the future. Unfortunately, since 2008, and out of necessity rather than intention, the public infrastructure industry has not been forecasting needs accordingly. This occurs whenever a “standard TCA ESL and replacement cost” is applied to repairable systems (i.e., civil engineered), which typically make up 95% of a municipal portfolio. In order to remedy this discrepancy it is necessary to properly apply the probability theory through engineering risk and reliability analysis, which is very similar to consumer-based forecasting in the private industry.

Through practical research at the University of Waterloo and collaboration from municipalities across Canada over the past five years, it has become evident that the following steps are necessary in order to reset the academic paradigm of public infrastructure asset management study:

1.  Analysis of all asset classes within an infrastructure portfolio

2.  Significant focus increase on forecasting rather than data collection

3. Application of consumer-based forecasting

The following section describes the necessary reset within the practical paradigm.     

4.        Practical Reset

In progressive societies, it is typical for industry to apply new technologies and methods developed in academia. This is to ensure that their benefits reach the public in a timely manner.  This section explains the practical application of the consumer-based forecasting approach for the provision of public infrastructure.

Table 1 shows the Standard Established Performance Measures (SEPMs) for which a municipality can collect data on. 

Table 1: Standard (non-age-based) SEPMs

Once collected, their future state is typically forecast according to a TCA schedule (i.e., age), which is not appropriate for 95% of a municipal portfolio as discussed in the previous section. However, this is not the only significant gap in current public infrastructure asset management practice. Equally and if not more important is the need to realize that SEPMs are not the only measures or factors which play a role in infrastructures’ objective performance. There are nine other equally important factors to decision making of when, what, and how to renew infrastructure.  These Corporate Decision Factors are listed in Table 2. 

Table 2: Corporate Decision Factors – Not Accounted for by SEPMs

[https://doi.org/10.1680/jinam.18.00039]

 It is SEPMs and corporate decision factors when analyzed together that provide the comprehensive or objective performance of public infrastructure. SEPMs by themselves are just one angle (i.e., engineering) of infrastructure’s performance, in practice there are another nine. To derive comprehensive or objective performance of infrastructure it is necessary to analyze past and current engineering, financial, and administrative municipal information.

The analysis uses engineering risk and reliability analysis to appropriately forecast future performance and corresponding expenditure of infrastructure. This is equivalent to the forecasting methods used in the private sector such as banking, insurance, and retail. The method results in “living” asset management plans as it allows for modelling performance according to corporate decision factors, not just SEPMs.

Preliminary research results indicate an approximate annual average reduction of 200% in projected expenditure associated with treating (repairing or renewing) poor performing portions of community’s infrastructure, over an analysis period of 25 years. This may appear surprising as it suggests that the current asset management paradigm is overestimating needs and therefore potential tax and user fee increases by a factor of two. However, considering the lack of scientific basis of prescribed TCA’s ESL for repairable systems that make up 95% of a municipal portfolio, this is to be expected. As such, it is imperative for future societal progress that forecasting of public infrastructure’s performance and corresponding expenditure be performed in a manner that is in tune with scientific principles as already applied in the private sector.

5.        Conclusion

In order to ensure continued societal progress it is necessary to reset the asset management paradigm as it applies to the provision of public infrastructure.

Asset management’s focus requires shifting from data collection to forecasting. From an academic point of view, within the field of engineering risk and reliability there is a clear separation of non-repairable and repairable systems. The former is primarily concerned with estimate service life and full cost replacement, while the latter is focused on mean time to treatment (repair) and average treatment (repair) cost.  Due to PSAB 1350’s implementation, currently non-repairable forecasting logic is being applied to repairable systems. Preliminary research results from the University of Waterloo indicate a 200% overestimation in infrastructure needs due to this misapplication of probability theory. Appropriate adjustments to PSAB 1350 are required in order to remedy this deficiency. This effort should be led by academia in collaboration with municipalities.

Through the necessary adjustment mentioned above, the challenge of quantifying objective infrastructure performance will be solved. Pure engineering metrics of infrastructure performance are not sufficient in quantifying its objective performance, as they do not account for corporate decision factors, which are a proxy of the consumer needs (i.e., community needs). As such, it is necessary to apply a consumer-based asset management approach in order to ensure optimal infrastructure decision are made. The essence of the approach is analyzing corporate information (engineering, financial, administrative) and ensuring repairable system’s forecasting logic is applied to repairable systems, which typically make up 95% of the municipal portfolio.