University CHP achieves efficiency

A university considered several design and upgrade alternatives for its central heating plant (CHP) to maximize energy efficiency and reduce emissions.

By Andre Pearson, PE, CEM, LEED AP, CBCP, and Jose R. Rodriguez, PEng December 29, 2014

Learning objectives:

  • Learn how to achieve maximum energy efficiency in existing central heating plants.
  • Understand advantages and disadvantages of lifecycle cost analysis of different alternatives in existing heating plants.
  • Understand how to reduce emissions in existing coal fire central heating plants.

Designers and managers of educational buildings continually face increased demands to improve the energy efficiency levels of their institutions. To reduce energy consumption, local and state energy codes, standards like ASHRAE Standard 90.1: Energy Standard for Buildings Except Low-Rise Residential Buildings, and guidelines like the U.S. Green Building Council LEED rating system have raised the minimum efficiency levels.

Shippensburg University (S.U.), Shippensburg, Pa., has recently implemented a “go green" initiative whose goals are reducing fossil fuel consumption and lowering energy costs throughout the campus. The university’s design and upgrade includes replacement of an aging central steam plant, which must satisfy both near-term and projected future space conditions.

Four options were devised for potential configurations for the utility systems. They compare technical and economic issues. This feature will review the characteristics of each option, economic analyses, and qualitative advantages and disadvantages.

According to a prior study, replacing the existing plant was not determined to result in a lower lifecycle cost. Moreover, additional regulations by the U.S. Environmental Protection Agency and Pennsylvania Dept. of Environmental Protection regarding air emissions of large coal-fired facilities could be expected within the lifecycle of a renovated coal-fired plant. These regulations would require costly upgrades to the emission control equipment. Current proposals by Maximum Achievable Control Technology (MACT) also require upgrades to address emissions of hazardous air pollutants. Environmental regulations are much less stringent for natural gas-fired equipment.

Central heating plant alternative designs

The existing central heating plant, shown in Figure 1, is approaching 60 years old. The large masonry (brick and concrete) structure is located near the main entrance to the campus, across from the historic portion of the campus that includes the president’s home and Old Main.

The central coal-fired heating plant generates steam that is distributed throughout the campus to heat the buildings from September through May. Each building uses a gas-fired summer boiler to meet heating requirements during the rest of the year.

The central heating plant is shut down in the summer months when minimal heating is required as it relates to keeping a large steam and condensate system in year-round operation.   

A stringent study (Table1) was conducted for the design of a new heating system that would be safe, reliable, sustainable, efficient, and cost-effective in replacing the aging central steam heating system. The new system also needed to satisfy both near-term and projected future space conditioning requirements of the campus.

An 8,760-hour simulation projected the heating loads of the campus for a typical year as well as how each option’s equipment would perform to meet the loads. The electric and natural gas energy uses of the systems were totaled to estimate their annual operating costs. Initial costs for equipment, distribution piping, installation, and plant construction were estimated for each option as well.

The campus steam demand typically peaked below 60,000 pounds per hour (PPH); however, plant operators reported that on rare occasions in the past, the steam load approached 70,000 PPH. Pipe distribution system losses were included in the steam load, and because of the poor condition of the piping distribution system, there were times where the distribution losses were significant. The higher steam demand was probably associated with the combination of higher losses from leaking pipes during periods of very cold weather. The steam demand in the dining halls also decreased when dedicated steam generators were installed to supply “clean steam” for kitchen use.
By reviewing steam records during periods with no space heating and little water heating, we estimate that, on average, the steam load associated with distribution losses was approximately 6,000 PPH. These losses were steady throughout the year, hour after hour, while the steam distribution system was energized. If steam or condensate leaks developed, distribution losses increased until the necessary repairs were made.

In the future, there will be added heat load as a result of buildings being added to the campus. However, there will a reduction in steam use through energy improvements in existing and replacement buildings, and reduced losses as a result of new steam distribution piping replacing the old pipe. Improved controls on major steam loads will also decrease the peak campus steam demand by scheduling the steam load to occur at different times. Consequently, the peak campus steam demand was expected to remain below 60,000 PPH in the foreseeable future. Over a longer period of time, the steam demand may eventually increase; perhaps as much as a 25% increase could occur, which would increase the peak to 75,000 PPH.

To verify the heating load for the campus, a peak heating demand for each building was estimated. When the building heating load for each building was added with the estimated distributed losses, the total load was approximately 62,000 MBh, which confirmed the estimates developed from the steam generation records described above.

Analysis alternative selected

Option 4 was selected because it provides the lowest initial cost and the lowest annual operating costs of the four, taking advantage of several key benefits:

  • High-efficiency condensing boilers, which save on annual natural gas costs
  • Minimal hot water distribution heat loss due to decentralized heating, which also reduces gas costs
  • Favorable project phasing with the minimal hot water distribution work, allowing existing steam line trenches to be reused for chilled water pipes, which saves on digging and installation.

Also, this configuration uses six sets of smaller hot water condensing boilers to supply low-temperature hot water (LTHW) to clusters of one to six buildings (near-term). Figures 2 and 3 show the Ceddia Union Building (CUB) and Reisner Buildings before and after construction.

The defining characteristic of Option 4 was a decentralized heating system, rather than a central heating plant. This would take advantage of the exceptionally high efficiency of condensing boilers, especially at part load with supply temperature reset based on outside air (OA) temperature, as well as eliminate much of the distribution pumping and heat loss of the other options. The boilers would be placed in a mechanical room of one of the buildings in each cluster, with piping running to the other buildings.

With this option, much of the campus’ existing utility infrastructure could be reused with appropriate project phasing. The existing steam plant could be converted, decommissioned, or repurposed, and the chilled water lines could be laid to reuse the existing steam pipe trenches as much as possible. This could save considerably on initial construction and digging costs.

The decentralized boiler plants in the near term would use between two and six boilers of 3,000-MBh capacity each. Table 2 describes the heating characteristics of each cluster.

The following are the sequences and assumptions that were used in the heating system model:

  • A hot water supply temperature reset schedule was developed, which adjusts the supply temperature based on OA conditions. The setpoint varies linearly with OA dry-bulb from 180 F supply temperature at 8.7 F OA (heating design spec) down to 130 F at 60 F OA, when the heating system would just be engaged.
  • A 20 F delta T across the loads was always assumed, which dictated the hot water return temperature.
  • Each cluster’s boilers were staged to run as many units as possible, each at the lowest possible loads but above their minimums (approximately 7% of capacity). Their combustion efficiencies are the best in this low part-load range (close to 99% efficiency).
  • Boiler efficiencies were determined from manufacturer-supplied curves, which relate the combustion efficiency to the return water temperature at different loads.
  • Combustion fan energy for these boilers has been neglected.
  • Each cluster is assumed to use one variable-speed hot water pump, plus one standby. The pumps modulate to maintain a set differential pressure setpoints in their distribution systems.
  • The required hot water flow (gpm) in each cluster was calculated for their given loads, assuming a 20 F delta T.
  • Each cluster’s maximum system head [ft WC] was estimating by totaling losses through each of the typical boiler and air handling unit (AHU) piping components. The system curve, which projects head as a function of flow (percent flow in this case) was then assumed to follow the same shape as the chilled water curve.
  • The power draw (kW) of each pump was calculated based on the flow and head in each hour and assuming fixed pump, motor, and variable frequency drive (VFD) efficiencies. Summing power draw across the year yielded the annual energy use of each pump (kWh/year).

Tables 3 and 4 summarize the nominal and operating characteristics of the natural gas and electric heating system equipment, respectively, including projected annual energy use and operating cost.


The overall economic summary for Option 4 is presented in Table 5, including all capital cost estimates. Trenching costs assume excavation of existing steam pipe trenches with less rock removal required, as compared with the other options. Construction of a new chiller plant building is assumed and includes bringing the necessary electric supply to the site.

The university’s primary focus is to provide a safe, reliable, cost-efficient, and environmentally friendly campus for students. By using natural gas on the project, S.U. takes advantage of a lower operational cost and is expected to save approximately $330,000 per year in electricity costs. Also, because the steam plant’s existing pipes were used for the upgrade, an additional $10 million was saved by not having to replace the underground steam system.

Most importantly, the team expects to reduce the campus carbon footprint by 31%.

Andre Pearson is lead engineer, infrastructure and optimization at WM Group. Pearson has 11 years of experience in HVAC design, energy audit, and commissioning services for various market sectors. He has completed numerous systems assessments/optimization projects and has identified millions of dollars in savings combined with improved reliability. Jose R. Rodriguez is a technical service engineer at Wallace Eannace Associates Inc. Rodriguez assists engineering consulting firms in New York City on the HVAC product segment of the company.