INTRODUCTION
You would think that energy efficiency is relatively simple: perform an energy audit, install the retrofits and then reap the energy savings. Unfortunately, it doesn”t always work that way. We performed an energy assessment of several stores of a major retail chain in the San Francisco Bay Area and identified a handful of low-cost retro-commissioning measures that had very promising potential. We quantified the expected savings and costs and returned after the project was installed. We then measured the savings using various methods and found either minimal or negative savings. The problem we discovered was that on nearly every measure, the contractors had repaired the hardware, but through various means had ensured that energy savings would not occur. This paper provides an account of the failed project at one store and the steps we took to remedy it. Specifically, this paper stresses the importance of Measurement and Verification and Commissioning of the retrofits.

DESCRIPTION OF THE BUILDING
The store, located in San Francisco, belongs to a well-known national retailer, whose name we will not divulge. The store is an aggregate of 3 buildings which have been joined together to comprise almost 1,000,000 square feet, of which over half is selling floor. Stock rooms and offices comprise the remainder of the space. The different buildings range between 8 and 11 stories tall.

The three buildings comprising the store were built at different times from the 1920s to the 1980s. Originally the buildings had different air handling, chilled water and hot water systems. Over the years, through energy conservation and facility improvement measures, the chilled water systems have been merged into one system.

There were no operating boilers in the store. Steam is provided to the store by an external vendor. Hot water is supplied to multi-zone air handling units and perimeter reheats in some areas of the store via heat exchangers.
There is one common cooling plant which houses two 500 ton centrifugal chillers (2004) which run all year. Chilled water is supplied to the Air Handling Units (AHUs) via primary/secondary chilled water loops. During the hottest months, both chillers run at around 90% full load””this happens about 5 days/yr. During the cooler months, one chiller runs at about 40% full load. If you have been to San Francisco you probably know that even in summer a typical day only reaches about 60 degrees . A properly designed and operating building in San Francisco should not need mechanical cooling most of the year, instead relying upon outside air to meet its cooling needs. This was obviously not the case .

A utility bill analysis identified an out of control building. Figure 1 presents twelve months of average usage per day versus average outdoor temperature. Each point represents a billing period. The superimposed red line represents the statistically insignificant trend. The lack of clear trend indicates that the building is either haphazardly controlled or that energy use varies due to some other variable. We believe mostly the former. During warmer periods (which are not that warm) the store uses more energy, indicating a variable cooling load based upon weather conditions. (An ideal system that uses outside air whenever possible should show a horizontal trend in this 48 to 66 degree temperature range.)

There are over fifty AHUs: a mixture of single zone, multi-zone, and variable air volume units. Each of the three sections contains different types of AHUs.
Electricity Costs for the store were over $2.5M per year. With the economic collapse in the fall of 2008, smart retailers were looking to cut costs wherever possible. One line item that could be cut was utilities. Saving 10% or more could add at least $250,000 to the bottom line.

BACKGROUND OF THE UTILITY PROGRAM
There may be several reasons why California uses less than 50% per capita of the energy than the rest of the country, but one major reason is the aggressive effort of the California Public Utilities Commission to cut energy usage. Commercial ratepayers of the investor owned utilities pay a fee in their utility bills that funds energy efficiency programs. These funds are then channeled to the investor owned utilities to promote energy efficiency. These utilities have over one hundred targeted programs aimed at different vertical markets such as: wineries, retail, hospitals, supermarkets, etc. Often these programs will include free energy audits or retro-commissioning services in conjunction with generous incentives to implement energy efficiency measures. In some cases, the utilities will pay for up to 100% of the cost for implementing the measures. The utilities administer some programs directly and outsource others. The outsourced programs are designed and administered by third party energy consultants.

Quantum Energy Services & Technologies, Inc. (QuEST), an energy consulting firm headquartered in Berkeley, administers a retail program for PG&E which covers the San Francisco Bay Area. This program offers retailers free retro-commissioning studies along with incentives to implement energy conservation measures found. The utilities give incentives to the building owners based upon the amount of energy saved. But in order for energy savings to be recognized by PG&E, these savings need to be measured and verified and then the savings calculations must pass a review by third party reviewers. Nobody gets paid if the work does not pass the third party review. The third party review process is necessary to prevent false claims of savings, or gaming of the system. The reviewers can be tough and require all assumptions to be documented and based upon published standards or guidelines. The drawback of third party review is that some measures are dropped as the Measurement and Verification (M&V) costs would be prohibitively expensive.

QuEST retained our company as a subcontractor to help out with the retail program. Our company performed Retro-Commissioning (RCx) services on 8 stores belonging to this unnamed retailer, and this paper is about one of the stores. However, the same story occurred at most of the stores. It wasn”t one failure, but many.

A NOTE ON THE LEVEL OF RCx RIGOR
RCx is different from energy auditing in that RCx typically involves a more detailed study of the building”s control systems and HVAC systems than energy audits. In addition, RCx typically focuses on repairing, recalibrating and reprogramming, rather than procuring new equipment. Simple paybacks for RCx projects typically are under 2 years. Examples of RCx measures are: repairing inoperable equipment, programming controls, demand control ventilation, and calibrating temperature sensors. Examples of energy audit measures (which are not considered RCx measures) are: installing energy efficient chillers, boilers or package units, converting single zone HVAC systems to variable air volume systems, and installing EMS systems. Energy audit measures often are more expensive and may have longer paybacks. On the other hand, true RCx studies are much more detailed, and thus much more expensive to conduct than energy audits. RCx studies usually involve data logging, functional testing of controls, operator training and post implementation commissioning which repeats much of the data logging and functional testing that was previously done. RCx is criticized by some as too heavy on the analysis, as it can require hundreds of hours of work just to perform the study, whereas energy audits consume much less labor.

In order to make the most efficient use of ratepayer dollars, in QuEST”s RCx program the amount of engineering time was scaled down to minimize the time spent on work that does not directly lead to energy savings. Rather than write commissioning plans, and 100-page Master List of Findings reports, the interim deliverable is instead an Excel workbook that describes the measure, states all assumptions and measured values, and calculates the savings. Equipment is data-logged or trended before and after the implementation of the measures. Calculations are made in Excel so they can be verified by third party reviewers. Written reports come later, but are less extensive than typical RCx reports.

ONSITE INVESTIGATION
Two engineers spent 3 days onsite examining the store”s mechanical systems, uncovering problems, and identifying RCx Measures. Our work to this point was nearly identical to an energy audit.
Once the RCx Measures were identified, the list of RCx Measures was given to the customer who then decided which of them should be pursued. The list also was approved by the third party reviewer.

MEASURES FOUND
We found the store could save about $300,000 in both RCx and Retrofit Measures, which, with incentives offered a simple payback of less than six months. That is 12% of their energy spend. The following measure types were identified and approved by all parties:

Retrofit Measures
1.Install Variable Speed Drives (VSDs) on Multi-Zone Air Handling Units (AHUs).
2.Installation of VSDs on secondary chilled water loops.
RCx Measures
1.Repair economizer control on some air handlers. Many outside air dampers were rusted in place. A two by six was used to prop one open.
2.Repair a small number of faulty VSDs, some of which were in bypass running at 100% fan speed.
3.Reconnect static pressure lines. Some VSDs were running at full speed because the lines running to the static pressure sensors in the ducting had been previously destroyed by contractors.
4.Repair/Replace stuck chilled water valves. These valves were cooling whether the AHUs called for cooling or not. As a result, sales floor temperatures ranged from 62 degrees to 70 degrees.
5.Connect some AHUs to the Energy Management System. These AHUs were running wild and had no control at all.

DATA LOGGING
Once the measures were selected by the customer, QuEST engineers placed data loggers to measure pre-implementation temperatures and power. Temperatures measured included Outside Air Temperature (OAT), Return Air Temperature (RAT), Mixed Air Temperature (MAT) and Supply Air Temperature (SAT). Fan Motor kW were also logged for those units on VSDs. Spot measurements were taken of Fan Motor kW for AHUs that were not on VSDs.

SAVINGS CALCULATION
Energy savings were estimated using bin data simulations. Like-type AHUs were combined. Special care was taken in calculating energy savings to ensure that savings were not double-counted. Each energy conservation measure was modeled assuming the prior measures were already implemented. We integrated the interval data that we collected into the bin data simulations. To do this, we created regressions of our variables (RAT, MAT, SAT, kW) versus OAT. These regressions were used to project RATs, MATs, SATs and kW for other outdoor air temperatures that were not included in our sample.

INSTALLATION
Once we had estimated savings using our bin simulation models and provided measure costs, the customer decided which measures to implement. They then hired contractors to implement the measures. VSDs were installed and repaired, economizer dampers repaired, AHUs connected to the EMS system, etc.

M&V PROVES NO SAVINGS
Once the implementation was completed, QuEST engineers returned to the site and again data logged the same temperatures and power as before. The resulting data, RATs, MATs, SATs and kWs, was again regressed against OAT. Using the regression, RATs, MATs, SATs, and kW values were again extrapolated and placed into the bin simulations.

The resulting calculations demonstrated the unthinkable. Not only were the energy conservation measures we had recommended not saving energy, the affected systems at the store were using more energy than before! Actually, this could be seen from just looking at the interval data. It was obvious that the economizers and variable speed drives were not working as intended. The “repaired” economizers were letting in less outside air than before, and the variable speed drives were still commanding the fans to run at a constant load, but at a higher speed than before.

QuEST alerted the customer that their investments were not saving energy. Facility personnel then investigated the problems, found them, and corrected them.

Even though the contractors had made the economizers operational (as opposed to frozen), the damper actuators were not calibrated correctly. When dampers needed to be fully open, they were not. When dampers needed to be at minimum position, they were not. The variable speed drives were also installed incorrectly. Some wiring and controls issues were resolved and the units started operating as expected. Once these issues were resolved, M&V was performed again. We repeated the data-logging and placed this information into our bin simulations, and again projected the annual savings.
There are many ways energy efficiency projects can go wrong.

“Faulty recommendations
“Poor implementation
“Untrained staff who compromise all the energy conservation measures undertaken

Faulty recommendations may arise from a lack of understanding of how systems operate or should operate. Years of experience, and a good understanding of physics and control theory is necessary to make sound recommendations.

Poor implementation has many causes, but often can be traced to the mindset that having the right equipment will make the difference. But as the lessons learned here illustrate, installing the right hardware is only half the solution. It needs to be integrated into the system and operate according to a logical and beneficial sequence of operations.

The last item is especially troublesome because it is so common. Even if the right hardware is installed and controls optimized, small changes to the sequence of operations made to “fix” local problems may have large consequences on overall system performance over time. Changing supply air temperatures at the air handler to resolve hot or cold complaints may upset the balance of the system and cause problems elsewhere. Professors at Texas A&M University have pointed out that in the absence of continuous monitoring, a building”s performance will fall to the level of the least-trained operator within two years.

HOW TO AVOID FAILED ENERGY EFFICIENCY PROJECTS
There are a couple of ways to avoid projects that fail to produce savings. After equipment is installed, it needs to be commissioned by a third party, not the contractor who implemented the ECMs. Commissioning can be expensive, but it is worth it. However, just because the equipment has been deemed operational by the commissioning agent, that doesn”t mean it is saving what was expected. Commissioning will tell you if the equipment is working as it should. To determine if you are actually saving what was expected, M&V needs to be done on the building. Although M&V can appear as a waste of money to some, it caught this disaster before it was too late.

Unfortunately, building owners often value engineer commissioning and M&V out of their projects and leave themselves open to big disappointments in their energy efficiency projects. M&V is like insurance””sure, it costs money up front, but the reassurance of knowing the project is done correctly should be worth far more than the initial outlay. What other product would you purchase without verifying that you actually received what you paid for? Why should energy efficiency be any different?

CONCLUSION
Unfortunately, energy efficiency isn”t as simple as we would wish. Energy consultants may deliver quality energy audits and RCx studies, but merely implementing sound energy efficiency recommendations does not guarantee energy savings. The weak link is often in the commissioning of the measures to ensure they are doing what they are intended to do.
To avoid underperforming on your energy efficiency measures, we suggest the following three strategies:

1. Commission what you implement with third-party commissioning experts. Commissioning agents are not interested in selling hardware. They are interested in making systems operate at peak performance. They understand physics and control theory and can identify and repair problems quickly.

2. Track your energy savings using M&V. Even using something as simple as utility bill tracking software can provide some insight into building performance. An increase in monthly energy usage when a decrease was expected would have triggered an investigation into the cause. Verifying performance at the system level (as we did), while more difficult and expensive, would have isolated the problem much more quickly and accurately.

3. Provide proper training so that your facility staff doesn”t override or bypass your energy efficiency projects. Although we barely treated this topic in this paper, this is probably the single most effective step you can take. Your staff is the brains behind building operation, despite what EMS vendors may say. Having the smartest control system will do no good if it is operated by the dumbest operators.