BEFORE THE ROOF CAVES IN II: APPA PU BL ISHED BY: RI CK BIEDENWEG AND RO BERT H UTSON T h i s a r t i c l e wa s or i gi n a l l y p u b l i sh ed by APPA (www.appa.org ) a n d wa s a u t h o re d by R i c k B i e d e n weg and Rob ert Hut s on wit h a s s i s ta nce fro m S ta n fo rd U n i ve rsi t y and t he Pacific Part ners Co n s u lt i n g G ro u p. Published By: Rick Biedenweg and Robert Hutson Before the Roof Caves in II: A Predictive Model for Physical Plant Renewal 1. Introduction and Context In 1980 a methodology was developed at Stanford University for predicting facilities renewal and replacement requirements based on life-­‐cycles and costs of building sub-­‐systems. The results were published by Rick Biedenweg and Robert Hutson in a paper entitled, “Before The Roof Caves In: A Predictive Model for Physical Plant Renewal.” Since 1980, Stanford has constructed new buildings, renovated a number of older facilities, responded to new code requirements, adapted to new academic directions, and developed a backlog of deferred maintenance. To assist Stanford University in an assessment of the current annual budget for future facility renewal, the Pacific Partners Consulting Group was retained to review the effectiveness of the original Before-­‐The-­‐Roof-­‐Caves-­‐in (BRCI) Model, incorporate new data, and to extend the predictive model to include the effects of adapting facilities to new program needs. 1.A Definitions The definitions used in this paper are drawn from Financial Planning Guidelines for Facility Renewal and Adaption, SCUP, 1989. This paper was a joint project of the Society for College and University Planning (SCUP), the National Association of College and University Business Officers (NACUBO), the Association of Physical Plant Administrators of Universities and Colleges (APPA) and Coopers and Lybrand. The project emphasized the need to recognize, plan, and budget for four distinctive aspects or cycles of facilities stewardship: on-­‐going maintenance; renewal; deferred maintenance or “catch-­‐up;” and adaption. While these four categories are linked (the definitions below discuss some of these links), they are also sufficiently distinct that the paper recommends treating them separately. On-­‐going Maintenance -­‐ -­‐ routine upkeep such as custodial services, unplugging drains, cleaning gutters, patching holes, fixing light sockets, lubrication of moving parts, and other routine maintenance duties that are required to keep the facilities operating. Failure to attend to these sorts of services results in accelerated deterioration of facilities and an accumulation of maintenance problems. Renewal or Facilities Renewal -­‐ -­‐ a systematic approach to repairing or replacing building subsystems such as roofs, HVAC, electrical, and plumbing systems, etc. to maintain and extend the life of the facility. Renewal is required to keep the building in good operating condition for its current purpose. This category of projects is sometimes referred to as Planned Maintenance. Page 1 Published By: Rick Biedenweg and Robert Hutson Deferred Maintenance -­‐-­‐ the accumulation of physical plant components in need of repair brought about by age, use, and damage from natural causes, and for which remedies have been postponed beyond the useful life of the system. Often, these corrections have been postponed due to insufficient funds. A continued underfunding for facilities renewal results in inadequate building renewal and increases the deferred maintenance backlog. Adaption -­‐-­‐ alterations in physical plant for changes in use, codes, and standards. These include code changes (such as ADA) and technology advancements (such as high speed networking) in addition to facilities that become obsolete for program reasons. The functions described above are closely related, but frequently planned for and financed through different management processes and budgets. However, decisions about timing and scope of projects in each of these categories may have significant budget implications for the others. For example, a major renovation of a facility to meet new health and safety code requirements might accommodate repairs or replacement of one or more subsystems not yet scheduled in the Renewal budget. While shifting the cost out of the Renewal program should affect that budget, it rarely does -­‐-­‐ usually because the Renewal budget is underfunded. 1.B Summary of the 1980 Sub-­‐system or BRCI Approach to Estimating Facilities Renewal Needs The sub-­‐system or Before-­‐the-­‐Roof-­‐Caves-­‐In (BRCI) approach to estimating future renewal and replacement requirements proceeds through the following steps: 1. Each building or facility is analyzed in terms of sub-­‐systems (those major components or systems such as electrical, elevators, roofs etc.) that have a significant impact on facility wear-­‐out and resulting replacement/renewal costs. The characteristics of sub-­‐systems are determined as follows: a. each sub-­‐system has an estimable useful life; b. sub-­‐system cost and life-­‐cycle data are available; c. the sum of the sub-­‐systems in a facility represent all of the components of the facility that could eventually wear out or require replacement. The sub-­‐system categories developed for Stanford facilities in 1980 included the following: Foundations, Major Vertical Floor, and Roof Structures; Roofing; Exterior Cladding; Interior Partitions; Interior Finishes; Elevators; Plumbing; HVAC – Moving HVAC -­‐ Static; Electrical -­‐ Moving; Electrical -­‐ Static; Fire Protection; Special Equipment; and Miscellaneous. 2. For each sub-­‐system an estimate of the life-­‐cycle is made. Page 2 Published By: Rick Biedenweg and Robert Hutson 3. Buildings with similar sub-­‐systems are grouped into categories. These categories include: High-­‐intensity Laboratories; Libraries, Offices, and Classrooms; Low-­‐intensity Laboratories; Athletics; Patient Care; Residential; and Miscellaneous. 4. Average renewal/replacement costs for each sub-­‐system are estimated in dollars per square foot for each category of facilities. 5. Facilities are classified into five-­‐year age cohorts1 by facilities type and by date of construction or most recent major renovation. For each cohort the total square footage of buildings is identified. 6. Projections are then developed for each cohort of facilities as to when specific sub-­‐systems would need replacement and how much it would cost to replace these sub-­‐systems. 7. These projections are then summed across all sub-­‐systems and facility categories to estimate total facility renewal needs per five-­‐year period. 8. Since the projections showed a highly cyclical pattern of expenditures, a “smoothing” or moving average was used. For more information about this approach see either Financial Planning Guidelines for Facility Renewal and Adaption, SCUP, December 1989, or “Before the Roof Caves In: A Predictive Model for Physical Plant Renewal,” APPA Newsletter, Association of Physical Plant Administrators, Part 1 (July, 1982), Part 2 (August, 1982). ___ 1 These age cohorts are identified by their midpoint. That is, the 1985 age cohort refers to the period from 1983 through 1987. Similarly, the 1960 age cohort refers to 1958 through 1962. Page 3 Published By: Rick Biedenweg and Robert Hutson 2. Benchmarks Using the 1980 BRCI Study To address the question of how well the 1980 study actually predicted facilities renewal needs at Stanford, the published results were compared with actual budgeted expenditures for facilities renewal at Stanford over the past ten years and deferred maintenance cost estimates prepared by an independent contractor. According to the basic theory of the model, the difference between actual expenditures made and facilities renewal needs (over any period of time) should be approximately equal to the increase in deferred maintenance needs over that same period of time. Figure 1, BRCI (1980) Deferred Maintenance Prediction, shows both the BRCI model results and the actual expenditure levels for 1985 through 1994, in 1994 dollars.2 The top line of the graph is the level of needed annual expenditures3 as predicted by the 1980 BRCI study. The line above the darkly shaded region represents the actual expenditures4 on facilities renewal (adjusted to 1994 dollars). The difference between these lines (the lightly shaded region) should represent the increase in deferred maintenance needs from 1984 to 1994. Figure 1 ___ 2 The index used to calculate 1994 dollars for Figure 1 is the inflation adjuster published in the Means Facilities Maintenance & Repair Cost Data 1995, page V-­‐17. 3 The predicted level of expenditures was taken from the “Before the Roof Caves In” paper published in Critical Issues in Facilities Management 4 -­‐ Capital Renewal and Deferred Maintenance, APPA, 1989 (page 28). 4 The large increase in 1993 was due to a $4.2 M special allocation for roofs. Page 4 Published By: Rick Biedenweg and Robert Hutson A cumulative increase in the expected deferred-­‐maintenance need can be calculated from this figure to be $48.8 million since 1984. For the purpose of testing this benchmark, it was assumed that there was very little deferred maintenance in 1984. This assumption was based on the substantial incremental funding made available in 1982, 1983, and 1984 -­‐-­‐ explicitly for the purpose of eliminating deferred maintenance needs of academic facilities -­‐-­‐ and from reports made to the Stanford Board of Trustees that deferred maintenance needs for these facilities had essentially been eliminated by the end of 1984. Thus, it was felt the $48.8 million could be used as a reasonable estimate of the expected total deferred maintenance need at the end of 1994. This $48.8 million was then compared with the actual deferred maintenance needs as determined by the independent study. To achieve accurate comparisons, however, certain categories needed to be excluded from the independent study. Those categories were foundations, structural systems, and exterior wall systems (BRCI excluded them from the analysis based on their extremely low failure rate) and floor coverings, interior walls, and special equipment (BRCI assumed these were the responsibility of the occupants of the facilities).5 After these exclusions, the figure from the independent survey was $47.9 million. While the similarity of these two figures is indeed remarkable, it is the belief of the authors of this paper that this degree of similarity (only a 2% difference) is probably accidental and that differences of ten to twenty percent are more like what should be expected. However, the similarity does support the reasonableness of the approach and further analysis, by sub-­‐system, was performed. This further analysis suggested a) modifications of certain sub-­‐system categories, b) the addition of new sub-­‐system categories, and c) an acknowledgment that facility obsolescence due to program reasons also needs to be considered. ___ 5 These sub-­‐systems accounted for $22.5 of the $70.1 million dollars of deferred maintenance needs on academic buildings reported by Dober/Lidsky. By sub-­‐systems they are: Foundations -­‐ $0.8M; Structural -­‐ $1.3M; Exterior Walls -­‐ $6.2 M; Floor Covering -­‐ $5.9M; Interior Walls -­‐ $4.4; and specialties -­‐ $3.9M. Page 5 Published By: Rick Biedenweg and Robert Hutson 3. Updated Model Results Using the new assumptions in the Before-­‐the-­‐Roof-­‐Caves-­‐In model, the expected facility renewal costs per five-­‐year period at Stanford were generated. These are illustrated in Figure 2. Figure 2 Overall, the model predicts renewal expenditures of between 0.6% and 2.7% of CRV will be needed on an annual basis, depending on the time period. However, a cyclical pattern and large swings from one five-­‐year period to the next are clearly evident in the figure.6 The large swings over time are due to the cyclical nature of both building construction and sub-­‐system life-­‐cycles. For example, the large 2015 period includes complete electrical and mechanical overhauls for all the facilities built in 1965. Since the 1965 cohort has more square feet of construction than any other period (except for major seismic renewal period of 1995) and because the largest components of the costs, by far, are the electrical and mechanical systems, this result is not surprising. Nor is the 2045 peak surprising -­‐-­‐ it includes all electrical and mechanical systems for the facilities constructed in 1995 (the largest period of construction/renovation in Stanford’s history). ____ 6 These large swings are one reason why it is impractical to use a fixed percentage of the current plant replacement value as a basis for budgeting for facilities renewal. Page 6 Published By: Rick Biedenweg and Robert Hutson 4. Obsolescence for Program Reasons 4.A Concept As defined in Financial Planning Guidelines for Facility Renewal and Adaption, plant adaption refers to alterations in a physical plant for changes in use, codes, and standards. These include code changes (such as ADA) and technology advancements (such as a need for high speed networking) in addition to facilities that become obsolete for program reasons. This discussion is limited to the latter category (facilities that become obsolete due to program changes) because code changes and technology advancements are extremely difficult to predict over long periods of time. In addition, only the obsolescence of high-­‐intensity laboratories is considered because: a) these facilities are the ones in which obsolescence is most evident; b) high-­‐intensity laboratories become obsolete more quickly than other types of facilities; and c) obsolescence in high-­‐intensity laboratories is more predictable and an obsolescence cycle can be estimated.7 Including adaption in high-­‐intensity laboratories in this update avoids significantly overestimating total facilities renewal requirements. A simple example illustrates this point. Consider a high-­‐intensity laboratory constructed in 1960 with a specific sub-­‐system that needs a $5 million major renovation every 30 years. The BRCI model would predict a $5 million expenditure in 1990, in 2020, in 2050, and every thirty years thereafter. If, however, this facility is a biotechnology building or some other high-­‐technology facility, then industrial experience indicates that sometime around 40 to 50 years after construction (i.e., the year 2000 to 2010) the facility will be obsolete for the program it supports and either the facility will require major renovation or it will be taken down and a new facility constructed in its place. If it is the case that the facility is completely renovated or replaced after 50 years (in the year 2010), then there will be no need for the $5 million sub-­‐system renovation in the year 2020 (the new or renovated facility will be only ten years old at that time, and it requires this sub-­‐system to be renovated only every 30 years). Thus, facilities that become obsolete due to program changes and require renovation or replacement at that time cause the BRCI model to overestimate facility renewal costs on future occasions. ____ 7 It can be argued that classrooms and low-­‐intensity laboratories also become obsolete for program reasons after some period of time. While obsolescence of these types of facilities certainly happens, no evidence was found of this happening on any systematic schedule. These categories of facilities have therefore been excluded from this “adaption” adjustment. Page 7 Published By: Rick Biedenweg and Robert Hutson 4.B Facility Renewal Predictions -­‐-­‐ Taking Adaption into Account Figure 3 shows the predicted facilities renewal costs both with and without taking into account a predictable obsolescence for program reasons. Note that the anticipated cost, is typically 10% to 15% lower when taking obsolescence for program reasons into account. Figure 3 Page 8