When I first set out to write this article it seemed a fairly straightforward topic. After all, ASHRAE and various other sources have researched and published equipment lifespan tables that give good advice on expected lifespans for typical equipment used in facilities. And if you want to base five- and 10-year capital plans on a generic table, then I guess it really is a simple process. But alas, the more you think about it, the more complicated and difficult it gets.

What I quickly realized is that there are multitudes of competing, conflicting, and, in many cases, ambiguous parameters that influence decisions regarding when the best time is to replace equipment. Add to this equation the idiosyncrasies of mission critical operations and the decision process can get even more (or sometimes less) difficult. Ultimately, the exercise is a financial analysis based on predictions, estimates, and best guesses. And when you consider that in many if not most cases the responsibility for performing these analyses fall upon a facilities manager instead of a finance wizard, the challenge gets even more daunting.

At a high level, consideration should be given to capital costs to purchase and replace assets, overall operating costs, expected maintenance costs, and the intangible costs of risk of mission impact. Basically the decisions on what, when, and how are based on a total cost of ownership (TCO) approach. So far, this is still fairly straightforward. But start looking at each of these aspects in detail, and the ambiguous parameters start creeping in. Let’s begin with capital costs associated with the purchase and replacement of a given asset.

By performing a quick market survey and making a few phone calls, it is fairly easy to get the purchase price for a “replace-in-kind” query for an equipment replacement. Make sure to include all the ancillaries, options, and supporting components that need to be replaced along with the asset, but most of this information is readily available. Also consider the costs to demo, install, and commission the equipment. For mission critical equipment, this analysis needs to take into account the “soft costs” such as working only during approved “maintenance windows” (typically non-business hours and therefore on overtime, nights, and weekends), and possibly temporary services and/or rental equipment.

But the problem here is the assumption that “replacing-in-kind” is the best decision. It most likely is for a simple fan or pump, but what about a centrifugal chiller? Does it make sense to stick with a constant speed, variable-vane water cooled chiller or would a variable speed, air-cooled chiller be a better choice? Should a waterside economizer capability be added to the mix? Now things start getting a little tricky because energy efficiency, maintenance costs, and capital costs must be added to the equation to make an informed decision. But that’s not all. What about the costs of hiring an engineering and design firm to select, specify, and produce construction documents for a system redesign required when the decision is made to not replace-in-kind?

Furthermore, there are the costs for new training of operating staff on the new system technologies, writing new procedures, and establishing new maintenance agreements, spare parts inventory, etc. And using the chiller example above, there may be other considerations such as reduced operating and maintenance costs associated with cooling towers that are no longer needed and reduced water chemical treatment costs since the condenser water open-loop system was eliminated. Some utilities and government agencies may offer financial incentives for replacing older less efficient equipment that obviously would not apply to the replace-in-kind solution.

Just trying to figure out the energy costs between the two solutions can be incredibly complex. First, the utility rate structure must be fully understood in order to determine the energy costs for each solution. Very rarely is using a fixed unit cost such as $0.10/kWh an accurate means of assessing energy costs. And even if it is an accurate assessment today, will it still be after utility rates are renegotiated in the future? Obviously the solution that uses the least energy will have the lowest energy costs. But the exercise is to quantify the actual savings to compare to all the other costs cited above. So understanding the utility rate structure and comparing it to the projected energy profile of the various solutions becomes necessary.

"  At a high level, consideration should be given to capital costs to purchase and replace assets, overall operating costs, expected maintenance costs, and the intangible costs of risk of mission impact. Basically the decisions on what, when, and how are based on a total cost of ownership (TCO) approach. So far, thisis still fairly straightforward. But start looking at each of these aspects in detail, and the ambiguous parameters start creeping in.  " 

Determining the actual cost of electricity can be difficult, but the cost savings for an incremental reduction in energy use is even more difficult. This is because most electrical rate structures are not a straight unit cost such as $0.10/kWh. Instead, electric rates are “structured” into rate components such as the “straight charge” ($0.10/kWh), an “excess charge,” maybe a “maximum demand charge” (aka the “ratchet” charge), or a “combination demand charge.” These rates can also be affected (adjusted) by “real-time” or “time-of-day” rates. Other factors can include “fuel adjustment charges” or adjustments based on the site’s power factor.

Again, if the overall performance of the two solutions being considered is equal, then some of these factors can be “normalized” or otherwise disregarded. On the other hand, we have all heard stories of a chiller failure during peak rate hours where the subsequent in-rush current from a restart causes a peak consumption value that “ratchets” up the electricity rate for the entire month. The resulting impact on the overall monthly energy bill can be significant. If the same chiller had a VFD and “soft start” capability the peak demand consumption impact would be avoided, so these factors need to be considered in the financial analysis.

There are other financial considerations that need to be taken into account such as depreciation, salvage (or resale) value, corporate tax implications, etc. Salvage costs can be a positive if the asset is in relatively decent condition and there is a “grey market” and potential reuse for the specific equipment being considered for replacement. Otherwise, there may be some value for sale as scrap. But if the equipment is contaminated or has hazardous materials such as asbestos, banned refrigerants, or even lead paint, then the salvage value could even be a negative in that you end up paying for proper removal, transportation, and disposal.

Maintenance costs are another important consideration not only in comparing the total cost of ownership of various solutions, but also in the timing. Many types of equipment, and especially large and complex equipment such as chillers, generators, boilers, and even some electrical gear (UPSs, switchgear, substations, etc.) should undergo what is typically referred to as “major maintenance” at predetermined intervals.

Using the chiller again for example, many consider normal maintenance to include cleaning the tubes in chillers annually or bi-annually, performing eddy-current inspection of chiller tubes bi-annually or tri-annually, and doing a major chiller overhaul every 10 years. The decision to replace an asset should include how these maintenance costs can be avoided if the decision is to replace, or included and even escalated if the decision is to defer replacement and increase the intensity and frequency of maintenance to offset reliability concerns.


The discussion above has focused on how to determine the  TCO and replacement costs of equipment. This entire exercise needs to happen concurrently with an equipment operating condition assessment program so expected remaining lifespan can be quantified. Generally speaking, there is rarely any justification to replace reliable equipment in good condition that performs as required and has the necessary capacity. And the inverse is equally true. Unreliable equipment in poor condition should be replaced before a failure occurs and when the replacement can be accomplished with advance planning and when alternative solutions vs. replace-in-kind can be evaluated, and implemented when justified.

An equipment condition assessment program should gather as much data as possible regarding equipment operating condition. This would include a review of the equipment reliability history, maintenance history, and a visual inspection at a minimum. Online condition monitoring data such as vibration analysis, thermography, and tribology (lubrication analysis) can also provide valuable insight into equipment health, especially when trended over time to identify progressive degradation.

As stated at the beginning of this article, the optimal time to replace equipment, and the decisions regarding whether to replace-in-kind or to upgrade or otherwise redesign the system, is typically a financial issue. So to add even more complexity, the predicted costs should be based on an economic analysis.

A quick and easy technique is the “simple payback” method. This method adds up the cost savings and other revenue projections and compares it to the required capital investment to determine how long it takes to recoup the investment.

Unfortunately, this ignores the cost of borrowing money, interest rates, inflation, changes in utility rates, depreciation, taxes, tax credits, and a whole host of other financial considerations. Ideally, the economic analysis would consider all these factors such as a “present-value/present worth” analysis, or other sophisticated analysis techniques.

I remember long ago when working for a large financial institution I requested funding for a capital improvement project. The project was justified primarily on an energy reduction and overall cost savings of 15% annually. My proposal was rejected since the company responded that they could take the same investment into the stock market and expect to make 18% annually. I resubmitted the same initiative, but instead justified it as improving the reliability of the central chiller plant for the mission critical data center that supported the company’s “market room.” The project was approved without further discussion.