Pipe Down

In its day—1906—the former Sears and Roebuck campus on Chicago's west side was a fine example of cutting-edge engineering, capable of generating its own power and heat and even hydraulic fluid for elevators. Over time, of course, the campus' central plant—despite a number of upgrades—passed its prime, particularly when the once-exclusive mail-order company shifted its fo...

02/01/2003


In its day—1906—the former Sears and Roebuck campus on Chicago's west side was a fine example of cutting-edge engineering, capable of generating its own power and heat and even hydraulic fluid for elevators. Over time, of course, the campus' central plant—despite a number of upgrades—passed its prime, particularly when the once-exclusive mail-order company shifted its focus to department store retail. That transition also saw the company relocate its offices to Sears Tower in 1970, moving again in recent years to suburban Chicago.

Amid all those changes, including new ownership, the campus became significantly smaller—half its original size—as a number of buildings were demolished due to lack of tenants. The remaining buildings now host a number of different tenants, including offices, cold storage for produce and even a health clinic. The most recent ownership change occurred in 2001, and a survey by the new owners, who manage the overall property, found the central plant's equipment in a poor state of maintenance and operating efficiency. In fact, operating the plant incurred a heating and cooling cost of about $10 per sq. ft.—exceedingly expensive compared to the more typical operational cost of about $1.50 per sq. ft. Contributing to these inefficiencies was the fact that the plant was simply oversized for what was left of the campus—a condition that had to be rectified.

What could be done?

When it came time to overhaul the M/E systems in the fall of 2001, the first task was a preliminary analysis of the buildings in order to determine the available options. The original campus, which served as Sears' main headquarters, testing laboratory and warehouse facility, was conceived as a standalone operation. It occupied almost 9 million sq. ft. of floor space, and was the largest office building in the United States at the time. The original central power plant was built with three large coal-fired high-pressure steam boilers to which five centrifugal chillers—with a total capacity of 4,500 tons—were added in the 1920s to provide campus cooling.

At the time, the entire campus was serviced by a plant providing chilled water, steam, domestic hot and cold water and water for the sprinkler system. The entire complex was served by an intricate underground tunnel system to enable pipe runs to feed steam and water to each building, and utility power was brought to the campus from a centrally-located substation. Three high-pressure boilers—producing about 65,000 lbs. per hr. of steam—were added to the central plant building in 1970.

In studying the situation, it was necessary to simulate the facility's energy use and evaluate multiple heating and cooling options. For the design team, major challenges included a lack of existing documentation for all building systems, the historical nature of the campus—it was recently added to the National Register of Historic Places—and a limited budget. The team, in conjunction with the owners, chose the most common-sense approach: to separate what was formerly the administration building from the power plant and provide its own heating and cooling plant.

A profile of the building indicated a peak-cooling load of 820 tons and a peak-heating load of 9.54 MBh. The profile also indicated that for 90% of the time, the building cooling load was less than 500 tons.

The design team recommended two chillers, one with a capacity of 500 tons and the other at 400 tons. Utilizing uneven-sized units in a plant is counterintuitive to engineers, who prefer two similar machines, but was deemed ideal in this case. Since most units have better full-load efficiencies—as opposed to part-load—chillers with different capacities were selected through accurate load modeling. Additionally, the primary chiller was selected based upon high part-load efficiency to maximize energy efficiency.

Selective siting

Keeping in mind the project's limited budget, and the fact that only part of the building was in regular use, cooling capacity was designed to meet basic needs with room for expansion in the near future.

The next challenge was to site the new plant. The criteria included:

  • Proximity to existing chilled water, steam and condensate mains to minimize the need for extensive piping modifications.

  • Space to house equipment with room for future plant expansion.

  • Architectural and structural modifications for housing equipment.

  • The ability to rig the equipment in its new home.

  • Cost and ease of maintenance.

Three possible locations were evaluated: on grade outside the administration building; the roof; and inside a smaller service facility behind the main building. The decision proved an easy one in that the tunnel with all the chilled water and steam piping ran below the service building.

Another deciding factor was that existing switchgear was also located in a basement below. New power lines could be routed to the service building with only minor modifications to the switchgear itself. The biggest issue was modifying the antiquated equipment to provide power while complying with current codes.

As far as the specifics of the new cooling plant, the following alternatives were considered: water-cooled centrifugal machines, water-cooled screw machines and air-cooled screw machines.

In the initial analysis, the air-cooled option was eliminated due to space constraints and significantly higher energy costs, which were in the range of 1 kW/ton. Water-cooled centrifugal and screw machines were evaluated as viable alternatives, but the water-cooled units would require the installation of cooling towers and pumps for heat rejection.

The final selection was determined based upon the best owning and operating cost. First cost for either the screw or the centrifugal machines was comparable. However, the centrifugal unit, if equipped with a variable-frequency drive, would have higher part-load efficiencies. The nominal part-load value for the centrifugal machine was 0.34 kW/ton, while that for the screw machine was 0.511 kW/ton. Since the plant operates at part loads the majority of the time, it was prudent to select a unit with higher part-load efficiencies to keep operating costs down.

Ultimately, the new plant was designed with a 500-ton centrifugal machine and a 500-ton cooling tower with space and connections to add 400 tons of cooling and an additional cooling tower cell at a later date.

It should be noted that the roof of the service building had to be structurally reinforced to support the new cooling equipment and the possible new tower.

A primary-secondary pumping arrangement for chilled water was designed with variable-frequency drives on the secondary pumps, since variable pumping above 10 horsepower can result in substantial energy savings with a quick payback. Pumps were sized for the entire cooling capacity of 900 tons. Furthermore, the old plant had a constant-volume system, so piping changes were required for the conversion to a variable-volume system. The cooling tower was also designed with a variable-speed fan, further reducing energy costs.

Heating to capacity

Unlike the cooling system, the heating component was designed for full capacity to ensure adequate heating, even in unused portions of the building. The two main challenges for the heating plant were converting the existing vacuum condensate system to a gravity system and finding a route to rig the boilers into an existing building.

Two different options were evaluated. The first was to use four small 100-hp scotch marine boilers, which could be rigged through existing openings in the building. The second option considered two 200-hp scotch marine design fire-tube boilers, with the need to widen existing openings. A cost analysis showed a clear preference for the second, as installation and energy savings far exceeded the cost of creating a larger opening.

Although packaged boilers were designed for the project, the installation required that the boiler shell be set in place and the burners, gas train and other accessories be field mounted. Apart from a lower first cost, the second option also offered lower energy and maintenance costs with a fewer number of boilers and parts.

Another challenge—converting the vacuum condensate-return system to a gravity system—was overcome by some creative piping changes in the tunnel under the building. The condensate-return system had to be split into a condensate-recovery and -return system with two sets of condensate pumps.

Determining feasibility

Retrofit projects such as these, involving upgrades to heating and cooling plants, can ultimately save owners substantial operating costs. But the first step in any plant upgrade is to commission a feasibility study evaluating existing conditions/operations and determining projected payback periods. The study must include costs for upgrading ancillary systems, which have a tendency of falling through the cracks. The study then becomes a major tool for the entire team and ensures the success of the project.

In the case of the former Sears facility, an 80-year-old plant was successfully upgraded with a payback of just eight months. Accurate building load modeling and equipment matching helped optimize the owning and operating costs. That, combined with technology advancements, resulted in a case study for considerable energy savings.



Results Say it All

Energy calculations showed the facility's new boilers to be substantially lower in energy costs. The new calculated annual building energy values were as follows:

Electric

On-Peak Consumption: 3,490,780 kWh

Off-Peak Consumption: 1,395,480 kWh

On-Peak Demand: 1,750 kW

Gas

Consumption: 60,420 therms

Total Energy Costs: $315,997

Total Building Area: 300,000 sq. ft.

To the above energy costs, maintenance expenses should be added to arrive at the total operating cost. Operating costs are estimated to be around $425,000 per year, which runs about $1.41 per sq. ft., as compared to the pre-retrofit cost of $10 per sq. ft.

The previous cost of heating and cooling the space was almost $3,000,000, whereas the cost of construction was about $2,000,000. Therefore, the project had a spectacular payback of eight months! Paybacks like these are not very common and some of the factors which helped improve the payback in this case were intrinsic to the project.



No comments