This article is authored by a team conducting ongoing research into the subject of maintenance staffing algorithms. It is adapted from their presentations at APPA's San Jose educational conference and annual meeting. Matt Adams is president of the Adams Consulting Group, Atlanta, Georgia, and writes the Facility Asset Management column for Facilities Manager.Joe Fisher is assistant vice president for facilities at Georgia State University, Atlanta, Georgia. Ted Weidner is director of facilities planning and management at Eastern Illinois University, Charleston, Illinois.

Every aspect of facility maintenance is coming under increasing scrutiny in these days of doing more with less. Administrators are calling upon the various departments within their schools and colleges to build up their resource budgets. Facilities professionals are finding themselves in a defensive position. It is difficult to explain exactly what the maintenance department does and what is required to support its functions.

Facilities management is complex. The staff required for this work must have a multitude of skills, and the work that the department performs is technical. What the lay business administrators see of the facility staff represents very little of their overall functions. Given this, the compilation of a built-up budget for maintenance must be concise and highly credible.

APPA and PGMS (the Professional Grounds Management Society) have made significant efforts to create the standards and guidelines that allow facilities administrators to create built-up resource standards for both custodial functions and landscape maintenance respectively. Unfortunately, little has been published in the way of standards for building maintenance. Research into the subject shows that most data available is either benchmark data (from a number of sources) or generic standards for commercial office buildings (from the International Facilities Management Association). No standards or frameworks exist for calculating the staff requirements for institutional building maintenance.

One possible reason for this may be the complexity of the maintenance function. A clear definition of maintenance and its components is not widespread at this point. Moreover, few attempts have been made to establish a baseline of maintenance staff trade definitions. In order to create standards for staffing, the basic staffing units must be defined. This dialogue is our attempt to represent the systematic development of maintenance staffing standards.

The basic framework and assumptions of this research are set forth in such a way as to allow simple interpretation by the institutional community. It is our goal to outline and present the most simple variables associated with staffing standards. As research continues, the basic building blocks of this work may be modified to reflect the conclusions drawn from the collected data.

Maintenance Definitions

Because the large majority of the campus constituency has no real understanding of the complexity of building maintenance, it is important from a modeling as well as a communications standpoint to clearly define maintenance. The variety of maintenance activities have distinct purposes and sometimes even varied funding sources. The unique nature of each drives the respective staffing needs. Maintenance is categorized systematically for this research.

Staffing model maintenance considerations, "True Maintenance":

Maintenance activities not considered:

In addition to the "maintenance activities not considered"are other activities considered routine in the maintenance department. However, the scope of this research is to create true maintenance standards. Estimation of the associated hours dedicated to ancillary work is ultimately a localized issue.

Maintenance Measurement

The measurement of maintenance is a critical element of staffing analysis. As APPA used the five levels of service in custodial service measurement, so must an empirical gauge be applied to the three forms of modeled maintenance. Previous efforts in this area have suggested that the maintenance measurement was based on the response time to customer requests.

However, response time is a metric that applies to a subset of the total maintenance task inventory. The more meaningful measurement of maintenance revolves around the timely completion of both the predictable/planned preventive and capital actions as well as timely completion of unplanned equipment failure repair/replacement.

Measurement for preventive maintenance is based on completion of the manufacturers specified procedures. The manufacturers specifications are comprehensive and represent the maximum preservation of equipment life cycle. One scenario might suggest that an institution sets a goal of 80 percent completion of all manufacturers preventive maintenance standards. The reduction in percentage comes not in canceling activities but reducing the frequency.

Planned capital maintenance is scheduled based on the life cycle of building components. Numerous sources provide recommended component life cycles, which may be used to schedule equipment replacement or overhaul. Adherence to this schedule and completion of unplanned capital maintenance in a timely fashion form a measurement. Percentage completion is appropriate here as well.

Repair maintenance is one activity that is gauged by response, but also in relation to the other activities. The total labor resource commitment to this activity is important in relation to the other activities.

Define Staffing

Staffing is "bodies on the job." For budgeting, bodies on the job are accounted for in FTEs, or full-time equivalents. This unit represents one or more working bodies supplying the equivalent of one year of work. It is here where the model encounters its first variable. What is one FTE? Each institution may offer differing benefit packages. There are variable vacation and sick days offered.

The maximum number of hours per year is 2,080. From this is subtracted vacation, sick days, and any other off-the-job time. Clearly this variability is so unique to each institution and even employee that it cannot be pinpointed. As such, the lowest common denominator for staffing is the working maintenance hour. A "maintenance hour" is one hour of work that occurs at the location of the respective equipment or project site. The total number of maintenance hours is convertible to FTEs at a local level. Existing staff and future hires have specific employment packages that are easily applied to a "pool" of required maintenance hours.

It is interesting to review the additional working time that is not part of the maintenance hour pool. This other time is a sort of "overhead." The overhead comprises a wide variety of elements, not all of which are consistent with productive work standards. Industrial engineering studies of trade staff at a variety of institutions have suggested that less than half of a given day is spent performing maintenance hours. The overhead of any maintenance department represents an opportunity for increased efficiency.

In this era of zero-based budgeting, it is clear that 50 percent overhead is difficult to justify. On the other hand, should the required maintenance hours exceed the total available staff FTE pool, a case is then made for increasing resources.

Trade Positions

If you examine the job descriptions you will find that there are many ways to define work and to split job responsibilities. This model uses crafts to build upon. Following is a list of basic job functions as they are most commonly referred to. It is important to note that the model only includes the immediate level of supervision (i.e., foreman) which is not listed but assumed; all higher levels are captured as administration costs so are not part of the model. Following is each functional group with associated skilled and other positions listed.

I. Preventive Maintenance
A. Skilled Trades

  1. Air Conditioning Mechanic—maintains and repairs heating, cooling, and air conditioning equipment and systems.
  2. Mason—maintains and repairs masonry and plaster systems.
  3. Controls Technician—maintains and repairs control instrumentation for lab equipment and systems such as HVAC, fire alarms, security, clocks, public address, sound, TV, satellite, gas detectors, pneumatics, and boilers.
  4. Electrician—maintains and repairs building electrical service systems and associated electrical devices. Is also capable of new installation.
  5. Elevator Mechanic—inspects, maintains, and repairs all types of conveying equipment per ANSI standards.
  6. Plumber—maintains and repairs building piping systems (gases and fluids) and associated fixtures.
  7. Skilled Mechanic—maintains and repairs hardware, lab equipment, mechanical equipment such as pumps, motors, compressors, fans, turbines, and heat exchangers. Also performs welding.

B. Other jobs

  1. Mechanic—performs maintenance and inspection tasks requiring basic skills such as oil machinery, change filters, replace belts, roof repair, ceiling system repairs, hardware repair (including doors), floor covering systems repair, exterior wall repair, insulation repair and replacement, and glass replacement.
  2. Painter—applies protective coatings to all types on interior and exterior systems.
  3. Trades Helper—person of limited mechanical skills that supports skilled positions or mechanics directly or through ancillary duties such as job site clean up and delivery of supplies.

II. Capital and Repair Maintenance
A. Skilled Trades

  1. Air Conditioning Mechanic, Mason, Controls Technician, Electrician, Plumber—same as above.
  2. Millwright—fabricates, welds, and installs metalwork.
  3. Pipefitter—possesses plumber skills and also fabricates high pressure piping systems (e.g., steam piping).
  4. Roofer—replaces all types of roofing systems per manufacturers specifications.
  5. Sheet Metal Worker—fabricates and installs all types of sheet metal structures for heating and cooling and in support of roofers.
  6. Carpenter—skilled in wood working activities such as building construction and cabinet making.

B. Other jobs

  1. Mechanic, Painter, and Trades Helper—same as above

III. Central Plant Operations
A. Skilled Trades

  1. Air Conditioning Mechanic, Controls Technician, Electrician, Pipefitter—same as above.
  2. High Voltage Electrician—possesses electrician skills and also operates and maintains high voltage distribution lines and equipment.
  3. Stationary Engineer—operates and maintains boilers, refrigeration compressors, and associated equipment.
  4. Skilled Mechanic—same as above.

B. Other

  1. Firefighter—supports stationary engineer by: pull fires, blow down boilers, check and record pressures and temperatures.
  2. Mechanic and Trades Helper—same as above.

There are several factors to consider when creating a maintenance program and identifying the annual costs. All building components requiring maintenance must be identified. Major building components, where they are discrete elements, are counted. Examples include doors, windows, chillers, boilers, pumps, motors, fan coil units, etc. Other building components that are more naturally continuous are measured in units such as linear or square feet. Examples of this include piping, wiring and cables, floor surfaces, walls, ceilings, and roofing.

Obtaining a complete and detailed list of all the components in a given building is extremely difficult in an older, constructed building. Detailed surveys and measurements are necessary to develop a list of components and require significant expertise to find those items that are often hidden by other components. It is difficult to determine the length of piping that is behind a wall and when the construction drawings are schematic. Obtaining this information from the contractor of new construction is preferable but still requires planning and some cost to notify the contractor that these detailed lists will be one of the deliverables of the project.

Once the components have been counted or measured, you must identify and separate the different types of components. Building components that function similarly may have very different maintenance needs; i.e., terrazzo flooring has quite different maintenance needs compared to carpet. Each of these items has a different useful life or life cycle. Terrazzo is typically considered to have a life greater than 50 years, while carpet is considered to have a life of less than ten years. In a simple life cycle cost analysis, an architect or engineer will use the first cost of a particular component versus the first cost of a similarly functioning component to determine the annual (or building life) cost for selection purposes. However, if one does not consider the annual maintenance (preventive and corrective) costs, the comparison will be flawed.

A third factor to consider is the type of maintenance and the qualifications required to perform the maintenance. This often sets the labor cost. In some states, these labor costs are not controlled by the facilities officer, and specific maintenance activities require very different maintenance personnel. Floor materials again illustrate this. Carpet can be installed or replaced by carpenters or a lower paid trade, while terrazzo requires a subset of the masonry trade that is specialized. It is also unlikely that a carpenter will be assigned the task of performing welds on regulated pressure vessels.

A fourth factor is the amount of time required to perform a given maintenance activity. Is it easier (shorter duration) to perform the preventive maintenance activity for one component type versus a functionally comparable component? Painting drywall may take less time than painting concrete block, but repairing damage resulting from errant equipment carts may not. In one case, two trades and some interactivity wait periods may be required, while in the second case the damage may be so minor that only one trade is needed to effect a similar result.

A fifth factor is the age of a component. The age is used to determine whether the component has remaining useful life or must be replaced. In some cases, major equipment should be overhauled as part of annual maintenance once a given age is reached. If component replacement is known it can also be used to reduce the annual maintenance costs and identify the one-time major maintenance cost. The shift in maintenance activity may not affect the overall O&M budget but may be used to justify expenditures regardless.

This discussion has identified most of the factors that can assist you in deriving the cost of maintenance for specific building components. However, the level of detail required also requires significant information, computational resources, and time to determine annual expenses. A more effective means to determine annual needs is to reduce the number of variables needed to make the expenditure predictions. This can be done by eliminating those variables that are dependent on other factors or which may be assumed to be homogeneous.

Likely variables to eliminate are the type of maintenance, the qualities of the person performing it, and the length of the maintenance activity. Regardless of the amount of maintenance to be performed on a given building, each building component must be attended by a specific individual who is trained in delivering the appropriate amount of maintenance. The service delivery is specified by the manufacturer and delivered by the trained employee. The employee is identified above in Trade Position. In an effort to identify the appropriate amount of maintenance, the quality of service delivery must be taken as an immutable factor. Therefore, the specific trade position can be factored into the building components as can the type of maintenance and the amount of time to deliver it.

The next variable that is logical to eliminate is the detail of building components. When you consider the number of different components that make up even the simplest of buildings, a means to eliminate this variable, with its many possible choices and complexities, will greatly simplify any problem.

It can be eliminated by recognizing that campuses have and continue to standardize buildings for maintenance and design. Similar buildings contribute to a memorable character or quality of the campus. Similar buildings also allow employees to be trained in fewer specialized maintenance functions and to work more efficiently. Similar buildings are not devoid of special or attractive architectural features, they utilize appropriate forms of standardization. Many APPA members have campus standards that specify the types and quality of building components to be installed.

Additionally, building codes and standards are increasingly prescribing the functional components for a building of a particular use. Laboratories require a set number of air changes to control chemical fumes. Offices will have similar architectural characteristics if only to avoid distinguishing differences of an egalitarian faculty. The federal government has standards for energy consumed by a facility, less than 50,000 Btu per gsf. So, one could then assume that for a specific facility use a particular group of building components must be present in order to operate the facility.

Elimination of the above variables leaves three independent variables that are considered in the proposed model. The remaining variables are area, use, and age. Building area may be used to normalize costs based on typical facility needs or space types. Instead of individual building components with three non-compatible units of measurement it should be possible to create a reasonably accurate model using only floor area. Colleges and universities utilize the NCES Postsecondary Education Facilities Inventory and Classification Manual (National Center for Education Statistics, 1994) to define, report, and compare how institutional space is used. Building age, or years since major replacement/rehabilitation, assists in identifying major maintenance activities that are not considered part of the annual maintenance budget. Use of this data as an input factor can make the determination of annual maintenance needs more uniform and without the need for expensive and detailed data.

Application of an area-, classification-, and age-based model to determine annual building maintenance needs then requires relatively generic data that may be maintained by the central institutional planning office. The ability to draw on data that may exist in a data warehouse eliminates the cost of duplicate or specialized information. It should also reduce the likelihood of errors resulting from non-communication of changes. Lastly, use of data that is known and understood by non-facilities administrators may improve the acceptance of the calculated expenditure needs or allow comparisons between on- and off-campus activities or programs.

This discussion offered in this article represents the initial thoughts of the authors in our attempt to develop quantifiable means for predicting annual building maintenance expenditures. The next steps will be to gather comments from other facilities managers on the approach to assure its validity. If inappropriate assumptions have been made, the proposed approach may be modified.

Data must then be gathered from as many sources as possible. Examination of the data may result in the creation of regional adjustment factors to compensate for differences in design standards or building codes as well as trade practices and wage differences. Some interpolation between the NCES classifications may also be required. Ultimately, a set of tables will be developed. It should be possible to create a program that will draw information about campus facilities from a data warehouse and the tables on facilities maintenance to quickly and easily provide recommendations for annual building maintenance expenditures.