The model is on a 483 kilobyte Excel 5.0 spreadsheet. All costs are in 1992 dollars.
A cogenerating baseload facility can supply most of the heat required in a district heating system (with peak heating supplied by heat-only boilers), and can use the balance of the thermal output in absorption chillers or steam-turbine-driven chillers.
Generally it is desirable to maximize generation of electricity (the premium form of energy) in cogeneration. In a steam turbine power plant or combined cycle plant, extraction of heat reduces electric output, with less reduction occurring with extraction of lower-temperature heat. In this analysis it is assumed that district heating distribution takes place via medium temperature district hot water (peak temperature of 250F) or its steam equivalent (15 psig).
Chilled water for district cooling can be produced with lower-temperature thermal energy through absorption chilling. In any specific situation a number of variables would need to be examined, including the value of electricity compared to thermal energy, the pressure of steam output from the cogeneration options being considered, capital costs for single effect vs. double effect chillers, and other factors. The energy efficiency and economics of optimizing the integration of cogeneration and district cooling is explored in a report recently published by the International Energy Agency.5
Whatever method is used to generate chilled water, large increments of peak-shaving chilled water or ice storage are far easier and more cost-effective to implement in a district system than in many dispersed sites. This can facilitate significant reductions in peak power demand compared to on-peak use of chillers.
For the gas-fired DES and the coal-fired DES, peak heating demands were 170 and 400 Million Btu/hour (MBtu/hr) and peak cooling demands were 12,240 and 28,800 tons of refrigeration (tons), respectively. These model cases were established as representative "building blocks" for the analysis, recognizing that DES exist in a wide range of sizes and types. To provide a sense of perspective, in Table 3 these peak heating loads are compared to peak loads for existing district heating systems. Heating utility systems (as opposed to the many college/university or other institutional systems) are indicated simply by the name of the city where the system is located.
Table 3 is far from comprehensive; as noted above, there are an estimated 5,800 community energy systems in the U.S. 20
| Table 3. Comparison of peak heating loads for model systems with peak loads in selected existing systems.25 | |
|
|
Peak heating demand
|
| New York |
13,548 |
| Philadelphia |
2,160 |
| Detroit |
1,800 |
| Indianapolis |
1,721 |
| Milwaukee |
1,200 |
| U. of Minnesota |
924 |
| Baltimore |
780 |
| Minneapolis |
744 |
| Lansing |
656 |
| St. Louis |
612 |
| U. of Michigan |
587 |
| Seattle |
540 |
| Denver |
498 |
| Cornell U. |
480 |
| Pennsylvania State U. |
444 |
| National Institute of Health |
402 |
| MODEL COAL-FIRED SYSTEM |
400 |
| U. of North Carolina |
384 |
| St. Paul |
375 |
| Hartford |
372 |
| Pittsburgh |
366 |
| Ohio State U. |
365 |
| San Francisco |
338 |
| Iowa State U. Ames |
300 |
| Birmingham |
276 |
| Lansing |
270 |
| Omaha |
260 |
| Harrisburg |
240 |
| Tulsa |
234 |
| U. of Colorado Boulder |
210 |
| U. of Oklahoma |
173 |
| MODEL GAS-FIRED SYSTEM |
170 |
| Oregon State U. |
156 |
| Nashville |
144 |
| U. of Alaska Fairbanks |
112 |
| Utah State U. |
83 |
| Fairmont |
50 |
| Willmar |
29 |
District energy systems benefit from the phenomenon of system diversity, i.e., the fact that not all buildings experience a peak demand at the same time. The "coincident" demand is the maximum demand on the district heating system. "Non-coincident" demand was calculated using a heating diversity factor (90%) based on district heating utility operating experience. This was used to estimate the amount of individual building boiler capacity required.
The percentage of heating derived from cogeneration (77%) and thermal-only boilers (23%) was based on the heating energy duration curve, with an adjustment to account for planned maintenance.
Distribution losses (5%) and the distribution pumping assumption (3 kWhe per Mbtu heating send-out) are based on utility operating experience with modern hot water district heating.
Non-coincident cooling demand was calculated with a cooling diversity factor (90%) based on utility operating experience.
The percentage of cooling derived from the various types of plant capacity (electric, absorption and storage) was based on the cooling energy duration curve.
Energy inputs to electric chillers were estimated assuming electric centrifugal chillers plus auxillairies (cooling tower, condenser pump and chilled water pump). Average annual efficiency was conservatively estimated to be 0.95 kWh/ton-hour.
For heat-driven cooling, single-effect absorption chillers were assumed, requiring 18,000 Btu/ton-hour in thermal energy input and 0.20 kWh/ton-hour electricity for auxilliaries. Chilled water storage was assumed to be charged primarily by the electric chillers using off-peak power.
Cooling distribution losses (4%) and distribution pumping (0.13 kWhe per ton-hour) are based on utility data. Combustion turbine emissions are based on natural gas-fired combustion turbine characteristics. Boiler emissions are from a recent study. 19
Based on the above assumptions, the energy inputs, emissions and annual outputs for the representative gas-fired DES were calculated.
Power plant emissions, boiler emissions and other parameters are based on the same sources as described above for gas-fired DES.
Power plant and boiler emissions are based on the same references cited for cogenerated DES.
Power plant emissions, building boiler emissions and other parameters based on the same sources as described above for the gas-fired conventional case.
The 1993 projections were used because they provide a breakdown of projected capacity for 1995-2000, whereas the 1994 projections did not. However, a revised analysis should be based on the projections in "Annual Energy Outlook 1996," which extend out to 2015 and include more gas-fired capacity and less coal-fired capacity. The 32 gigawatts assumed in this calculation is 13% of the new generating capacity the EIA is now projecting for the years 1995-2015.
Only coal-fired plants and combined cycle gas-fired plants were targeted for this analysis, although other types of plants (e.g., renewable biomass plants) can also be used in DES.
Based on the energy and environmental characteristics of the DES cases compared to the corresponding conventional cases, the total reductions in primary energy consumption and air emissions by 2010 were calculated.
Supplemental boiler capacity costs were based on costs for a 30 MBtu per hour boiler plant developed by the Minneapolis Energy Center, including construction of plant building and all installed costs, inflated by the Producer Price Index.
Chiller plant capital costs were based on past and projected costs at District Cooling St. Paul, including plant building space and all auxilliaries.
Distribution capital costs were calculated after deducting estimates of the amounts of existing DES distribution capacity with which new cogeneration might be integrated. Estimates of existing distribution capacity were based on district heating and cooling production capacity data. Surveyed hot water district heating systems totalled 10,535 MBtu per hour in production capacity, and surveyed district cooling systems totalled over 1,530,000 tons of cooling capacity.1 Surveyed systems were estimated to account for 15% of total national production capacity, from which over 70,000 MBtu per hour hot water district heating capacity and over 10,000,000 tons of district cooling capacity can be inferred. In the present analysis, it was assumed that 35,000 MBtu per hour and 5,000,000 tons of district heating and cooling distribution capacity, respectively, could be used in conjunction with new cogeneration capacity, or half of the total existing capacity estimated per the methodology just outlined.
District heating distribution capital costs were based on actual costs to construct a complete new hot water district heating distribution system in St. Paul, Minnesota, inflated per the Producer Price Index. District cooling distribution capital costs were based on projected distribution costs for the St. Paul district cooling system when system size reaches 18,000 tons.
Building interface costs for district heating were based on interface experience in St. Paul with hot water heating. These connections are always indirect (i.e., require a heat exchanger). Building interface costs for district cooling were based on an average based on 12 studies of potential district cooling customers. This reflects a mix of direct and indirect connections. Most building connections are direct but some require a heat exchanger due to hydraulic pressure considerations.
Individual chiller capacity must be sized to handle a building's peak demand (noncoincident peak), plus a safety margin. Experience in Minneapolis and St. Paul with buildings which have converted to district cooling shows that the actual building peak demand is generally 60-80% of the previously installed chiller capacity. Assumptions were made regarding the percentages of the avoided chiller capacity represented by retrofit of existing chillers (20%), replacement of existing chillers (40%) and installation of complete new chiller systems including auxilliaries (40%). Costs per ton of refrigeration were based on analysis of chiller retrofit/replacement studies for buildings in St. Paul.
Peak heating demand served by DES was derived from previous calculations, and 40% of this demand was assumed to be attributable to new buildings (therefore avoiding building boiler capital costs). A safety margin was applied to calculate total new boiler capacity, consistent with design practice. (Actual peak was assumed to be 75% of the boiler capacity.) Capital cost per MBtu/hour was based on a recent study.19
Operation and maintenance costs are based on district energy utility operating experience and other sources, adjusted to reflect sharing of cogenera-tion operation and maintenance (O&M) costs. Electric O&M costs are based on Table 12 of the Climate Change Action Plan Technical Supplement.6 General and administrative costs were based on industry statistics.
Debt service was calculated assuming 7.5% interest over 20 years, and total annual costs were determined, expressed in billions of dollars per year and as costs per unit of end-use energy provided.
The prices of boiler fuels are Energy Information Administration data for commercial sector natural gas and distillate oil prices. Boiler O&M includes $0.89/MBtu for regular preventive maintenance and $0.18/MBtu for annual allowance for major maintenance, and $0.50/MBtu for operational labor.
Chiller O&M includes: $0.021 per ton-hour for maintenance, $0.016 per ton-hour for cooling tower make-up water and water treatment (3 gallons make-up water at $4.25/1000 gallons for the water and $1.00/1000 gallons for water treatment); and $0.03 per ton-hour for operational labor.
Debt service was calculated over a 20 year term, with capital for power plants charged at 7.5% interest, and capital for building equipment charged at 9.5% interest.
Annual costs of conventional approaches were then totalled and compared to totals for DES.
A scenario was developed for implementation of coal-fired and gas-fired DES, expressed in MWe of power plant capacity provided or avoided, consistent with the 1995-2000 and 2000-2010 projections discussed earlier. Annual capital investments for the DES capacity were then allocated, with the capital costs in these annual "front-loaded," i.e., unit costs were increased in early years and decreased in later years to account for the nature of investments in DES, in which early-year investments are significantly higher than long-run capital costs per unit of capacity.
A scenario for coal-fired and gas-fired conventional power plant capacity costs and building systems was then developed, consistent with the scenario for coal-fired and gas-fired DES.
Differences between the DES scenario and the conventional scenario, in fuel consumption, emissions and costs, were then calculated.