Energy Management Case Studies

St. Paul's Hospital Lighting Retrofit

Table of Contents

St. Paul's Lighting Retrofit

St. Paul’s Hospital is a 200-bed acute care facility located in Saskatoon, Saskatchewan. The facility is made up of three interconnected buildings. The present hospital was completed in 1963. A major addition was constructed in 1988. A nurses’ residence, which was built in 1928 is currently used for additional office space. All three of these buildings are steam heated from a central boiler plant. The facility is cooled by chilled water from a central chilled water plant.

As part of an ongoing energy efficiency program St. Paul’s Hospital initiated a major lighting retrofit. Lighting retrofit projects have long been identified as a simple energy conservation measure that results in substantial decreases in electrical demand and consumption. Lighting typically accounts for 40% to 80% of facilities power costs. The St. Paul’s Hospital retrofit project encompassed the changing of 1500 fluorescent light fixtures. The project started in the spring of 1999 and was completed in the summer of 2000. The majority of existing fixtures were standard 2 or 4 lamp fixtures housed in a standard 4-foot ceiling tile grid. In addition to the replacement of magnetic ballasts with electronic ballasts and replacement of the standard 34 or 40-Watt T-12 fluorescent tubes with two 30-Watt T-8 fluorescent tubes, aluminum reflectors were also installed. Tests showed that equivalent or better light levels could be achieved with the installation of an aluminum reflector with one 30-watt T8 fluorescent tube. Without the reflector two 30-watt T8 fluorescent tubes were required.

The case study provides an explanation of the work necessary to complete the project, outlines the background work, actual retrofit installation and the results of the project. It does not contain detailed technical information regarding the light output of the fixtures, etc. The use of reflectors as part of the retrofit was carefully considered and tests undertaken to verify performance did not negatively affect the needs of staff and patients.

The total cost for the lighting retrofit including labour was $85,000. The estimated electrical savings were projected to be $28,000 per year, which provides the hospital with an attractive investment payback of 3 years.


The study showed that a typical 2 lamp 4-foot fluorescent light fixture with magnetic ballast consumed 81 to 96 watts depending on the bulbs and ballast installed. A bulb for bulb replacement program in which an electronic ballast and two T-8 lamps were installed would decrease the power consumption of these fixtures to about 65 watts. With the implementation of the lighting retrofit program the power savings for St. Paul’s Hospital was projected to be between 16 and 31 watts per fixture. At this stage of the process, however, the persistent lobbying of a maintenance staff member resulted in the team exploring the option of using light reflectors. The option specified that the existing fixtures be modified with electronic ballast, single T-8 tube, and an aluminum reflector. Using this combination of components, the power consumption was projected to decrease by an additional 35 watts, resulting in a power saving of somewhere between 46 and 61 watts per fixture.

The installed cost for the electronic ballast, single T-8 tube, and aluminum reflector was double the cost of electronic ballast and two T-8 tubes. The economic analysis showed that with the reflector installation the payback was only 18 months wherever the fixture was on 24 hours per day.

The fixture modification was as follows:

  • Replace the magnetic ballast with electronic ballast.
  • Remove existing lamp sockets and bracket holders for the two tubes and install new lamp holders and sockets for the single tube. These new lamp holders were installed to position the single tube down the middle of the fixture.
  • Install a highly polished aluminium reflector that serves as a new ballast cover, as well as the device that reflects a large portion of the light.
  • Install the new T-8 lamp in the new lamp holders.
  • Replace the existing plastic lens in the fixture.

The reflector kits contained all the hardware necessary to complete the fixture retrofits. Even the existing serviceable lamp sockets were replaced so that all the fixtures contained new internal components after the retrofit.

All fixtures upgraded in this project were ones that were on continuously. This decision was made for several reasons including: 1) 24-hour fixtures provided the most attractive payback, 2) as the retrofit crew gained experience and the installation process was streamlined, savings in labour costs would be realised making several future borderline lighting initiatives viable, and, 3) most of the targeted 24 hour lighting fixtures were located in corridors and other public spaces deemed as easy access areas for the retrofit crew. This provided a suitable environment for experimentation and the establishment of the protocols, which were followed for the project.

Audit Retrofit Plan

Once the protocol for retrofitting the lighting fixtures with an electronic ballast, single T-8 tube, and highly polished aluminium reflector was established another concern needed to be addressed. The perception of those using the facility was of paramount concern. It was necessary to demonstrate that the new fixtures would not negatively affect the lighting levels of the facility? More subjectively, how would the staff and clients perceive the new lighting? Would their satisfaction levels be compromised? It was also important to make sure that lighting levels remained constant from one area of the hospital to the other. Another issue, was the impact of colour rendering in a hospital environment? Visible change in skin colour is a symptom commonly used to monitor a patient’s condition. It is critical therefore that lighting levels are uniform to avoid the potential for false diagnosis due to apparent changes in skin colour caused by lighting.

Large arrays of energy conservation lighting fixture options are available on the market. While the products on the market appear to be similar the supporting technical literature is diverse and complex. Terminology describing the colour rendering index, photometric data, and reflectivity characteristics were some of the performance terms promoted in the product literature. Notwithstanding the abundance of information it was difficult to discern how the electronic ballast, single T-8 tube, and aluminium reflectors would perform particularly in relation to the potential impact on staff and patients as discussed earlier.

A trial and error approach was implemented to test performance of fixtures with aluminum reflectors. Two fixtures were retrofitted with reflector kits, and installed in a service corridor. The initial study results were very positive. The staff and patients were unaware of differences in the lighting performance between the current and modified fixtures.

A more scientific evaluation was undertaken to verify the feedback from staff and patients. The protocol required that light level measurements be taken at several corridor locations. These measurements were taken before and after the retrofit. The results mirrored the survey results and showed that lighting intensity (lux) was 10% greater after the new ballast, single tube and reflector kits were installed. This modest increase in lighting levels was deemed not to be a problem, particularly in view of the fact that light levels tend to degrade as fluorescent tubes get older. Based on these results the electronic ballast, single T-8 tube, and aluminium reflector retrofit was implemented.

CSA Approval

The proposed retrofit protocol required that the old bulb holders and metal ballast covers be removed. The new electronic ballast was then located to the side of the fixture and the reflector installed. Finally, a new holding bracket positioned the T-8 fluorescent tube in the middle of the fixture. The changes to the existing fixtures would result in the Canadian Standard Association (CSA) certification being withdrawn. Hospital policy mandated that all fixtures must meet and be certified to the designated governing standard. In the case of lighting fixtures CSA was the standards writing body. CSA allow modified lighting fixtures to be re-certified. The proponent can choose one of three options to participate in the CSA program. The first option requires that each modified fixture type be shipped to the CSA laboratory for testing. The second option is to have CSA representatives inspect installed retrofit fixtures. The third option allows CSA representatives to inspect each type of modified fixtures at the hospital site prior to installation. For economic and convenience reasons the third option was chosen. Twelve sample fixtures, each with different configurations, were prepared for inspection by CSA. Upgrades specified by CSA were implemented and all samples received CSA approval. The hospital was authorised to affix regulated CSA labels to any fixture, which conformed to the samples tested at the hospital site and approved by CSA. The authorisation requires the hospital to track all installations and issue a file number for each fixture type. CSA reserves the right to conduct unannounced visits to the hospital to verify labelled fixtures and records conform to CSA rules and regulations.

Retrofit Program

The project was completed using “in house” staff. The retrofit crew consisted of two people. The project coordinator was responsible for ordering all the material, tracking the project progress, and assisting the electrician assigned to the project.

The majority of the work was located in hallways with 8-foot ceilings. The retrofit team customized a rolling scaffolding that was used to access the lighting fixtures, serve as a mobile workbench and stock cart. Once the retrofit protocols were established and the crew was operational, between 25 and 30 fixtures were completed per each 8-hour shift. This allowed time for start up in the morning, clean up, breaks, and unexpected interruptions.

The cost of single tube ballasts was more expensive than two tube ballasts. The manufacturer’s wiring diagram, however, showed that the two tube ballasts could be used to power a single tube. The manufacturer verified that the use of the two-tube ballast with one fluorescent tube was acceptable, and would not harm the ballast though the life of the fluorescent tube would be reduced by 5%-10%. This negative was balanced by the fact that the light output of a single tube powered by two-tube ballast was greater than one powered by the single tube ballast. The new light fixture output increase (measured at 10%), combined with the fixture cost reduction was considered to be more of a benefit than the drawback of reduced tube life.

Tests revealed that light output decreased the deeper the reflectors were mounted in the fixture. The closer the fixture was located to the lens the greater the light levels. In one test location the light output of a reflector mounted 3" deep in a fixture was measured to be 15% less than for a fixture with the reflector mounted ½" into the fixture. To maximise light output the reflectors were placed as close to the lens as possible

Standard reflectors were available on the market but each manufacturer designed their reflectors with unique internal dimensions. The manufacturer chosen to supply reflectors custom-made “exact fit” reflectors and was found to be less expensive than competitors who offered "off the shelf" standard size reflectors.

Other areas of the hospital facility were included in the search for potential energy savings. Even though large windows in the hospital cafeteria provided excellent natural lighting during the day fluorescent lighting remained on. A photocell was installed to automatically switch off the fluorescent lighting circuits during daylight hours. Also a time clock was installed to automatically turn the lights off during the night when the cafeteria area is not commonly used. To provide flexibility and accommodate unexpected circumstances, switches were installed to enable occupants to override the time clock and photocell. This initiative reduced the “on time” of 80 two-lamp fixtures from 15 hours to 5 hours per day.

The hospital shipping and receiving area was examined. The area had approximately 200 two-lamp fixtures installed end to end as row lighting. The magnetic ballasts and T-12 fluorescent lamps were replaced with electronic ballasts and T-8 fluorescent tubes. No reflectors were used in these fixtures.

Due to the row lighting configuration it was possible to maintain more then adequate lighting for this area by only installing about half as many bulbs as were removed. Since the lights were installed in long rows, multiple fixtures were wired to be fed from a single ballast. The two lamp magnetic ballast in each fixture was removed, and a single 4 lamp electronic ballast was installed to feed a single tube in each of four fixtures in a row. The fixtures that contained these ballasts were marked with a small sticker for ease of identification for future maintenance. This greatly reduced the cost and number of ballasts that would have been required if single or two-tube ballasts were used.

Attention was then focussed on the stairwells where testing of the lighting output revealed an oversupply of fixtures. Approximately half of the existing fixtures were disconnected. When removing excess lighting, the removal of fluorescent tubes is not an efficient delamping strategy since magnetic ballasts will continue to draw power even when a tube is removed. To maximize energy savings the magnetic ballast on the fixtures left in service were replaced electronic ballasts, and T-8 fluorescent tubes installed.

Exit signs were also identified as another opportunity to save energy. The City of Saskatoon Fire Department has approved several types of phosphorescent exit signs for use in the marking of exits of public facilities. The regulations for use of these exit signs include: all phosphorescent exit signs must be installed near a light with emergency power backup. They shall not be installed in areas, such as equipment rooms, where lights may be off for extended periods. Phosphorescent exit signs are an attractive energy and safety option. Once installed these exit signs require no maintenance and consume no power. Since they require no electrical connections, additional signs can be easily added to better mark building exit routes.

Reflectors were not always a cost-effective option. This was the case at several corridor locations where existing lighting fixtures with 2-foot by 2-foot fixtures loaded with four 2-foot tubes and magnetic ballast were analysed. These fixtures were retrofitted with two 2-foot tubes and electronic ballast. The tests revealed that adequate light was provided and reflector enhancement was unnecessary.

Methodology to Estimate Savings

The lighting circuits in the hospital facility are not separately metered. It is therefore difficult to determine the actual final savings attained by this retrofit. The electrical loads placed on the facility are constantly changing through addition or removal of computers, medical equipment, or other building power needs. Concurrent with the lighting retrofit the power factor correction equipment was upgraded. Due to the variable consumption loads, there is no method to separate demand savings obtained through the lighting retrofit from those resulting from improvements to the power factor correction equipment. Instead individual fixture power consumption measurements were used to estimate the before and after performance of the retrofit fixtures

Savings Achieved

The following summarises the quantity, location, description and savings of each retrofit.

  • 900 2-lamp fixtures had reflector kits, electronic ballasts and single T-8 tubes installed. This combination resulted in a conservative average reduction in power consumption from 81 watts to 35 watts per fixture. These fixtures were all on 24 hours per day, for a yearly estimated power reduction of 362,664 kWh. The associated demand reduction was estimated to be 37 kVA. 200 4-lamp fixtures had electronic ballasts and 2 T-8 tubes installed. This combination resulted in an average reduction in power consumption from 162 watts to 60 watts. These fixtures were on 24 hours per day, for a yearly estimated power reduction of 178,704 kWh. The associated demand reduction in this case was estimated to be 20 kVA.
  • 350 2-lamp fixtures had electronic ballasts and two T-8 tubes installed. This retrofit resulted in a conservative average reduction of power consumption from 81 watts to 60 Watts. These fixtures are on 24 hours per day for a total yearly estimated power reduction of 64,386 kWh. The associated demand reduction was estimated to be 7 kVA.


The three phases of the lighting retrofit account for an estimated total power saving of 605,750 kWh per year, and a demand saving of 64 kVA per month. At the time this work was completed, the cost for power was $0.0291 per kWh, and $11.75 per kVA per month. This translates to a total estimated saving of $26,650 per year.

Additional miscellaneous lighting retrofits such as the timer and photocell in the cafeteria; the exit lights; and as well the conversion of incandescent bulbs to compact fluorescent bulbs resulted in an estimated additional savings of $1300 per year.

The labour and materials costs for the retrofit was approximately $85,000. The estimated savings for the retrofit are $28,000 per year. The overall project pay back is calculated to be 3 years.

In addition to the economic savings, the facility now has 1500 fixtures with new fluorescent tubes installed. The fluorescent tubes have an estimated life span of 20,000 hours, which should provide approximately 3 years of service before any of these bulbs have to be changed. In the short term, this will reduce the number of maintenance hours spent changing lights.

The retrofit exercise has provided the hospital with an opportunity to monitor the performance of the fixtures. As an efficiency measure all fluorescent tubes will be replaced at the end of their expected life. This will reduce the labour intensive process of replacing fluorescent tubes as they burn out.

Recommendations for Project Replication

In the two years since the lighting retrofit project was initiated savings have been achieved and feedback has been positive. Requests by other departments and groups to have their lighting systems upgraded have already been initiated.

Initial concerns that reflector performance would be affected by build up of grime and dirt have not been realised. Several of the fixtures in service for two years exhibit no visible build-up. It is anticipated that the reflectors will only require wiping when fluorescent tubes are replaced.

Concurrent with this lighting retrofit, six long-term care facilities in Saskatoon participated in a lighting retrofit program modelled on the St. Paul's lighting retrofit. The same crew performed the retrofit work.

In all cases, the use of the reflectors in these institutional facilities resulted in energy savings and improved lighting. The facilities were older and staff, patients, residents and the general public have welcomed the "shedding of new light" on old corridors.

For more information regarding the St. Paul's Lighting retrofit program contact Darrell Solie P.Eng @ 655-5019