Printer Friendly

Combined heating and power using microturbines in a major Urban hotel.


Annually, commercial buildings in the United States use about 4.2 billion MWh (14 quadrillion Btu) of energy in the form of electricity. Their electricity consumption is projected to grow by more than 50% by 2030 (DOE 2007). Because the electricity is generated at remote power plants, about 70% of this energy is lost in the conversion and delivery process, so only 4 quads of electricity is actually delivered to the buildings. Combined heat and power (CHP), on the other hand, is able to produce electricity and useful thermal energy so that less than 30% of the available energy is lost (Petrov et al. 2006). The cogeneration of heat and power is not a new technology. In fact, the highly efficient CHP process was used in many of the earliest electric generating plants, including the 770 MW Waterside plant built in Brooklyn in the 1920s (ConEdison 2007). Further, since the thermal energy stream is produced from recycled generator exhaust, the total amount of pollutants is less than if the energy were provided using a conventional combination of grid-supplied electricity and local boilers. Some observers have contended that small CHP units displace only highly efficient central gas-fired combined-cycle power plants and that the total emissions may therefore increase. However, a study directed at this issue found that both the total emissions and the total cost to the economy were reduced for every CHP scenario considered (Hadley et al. 2003).

A review of the national inventory of CHP installations as of 2004 found that most are located in industrial applications, such as paper mills and chemical plants (RDC 2005). This same review found that CHP units with electrical capacity between 80 kW and 1 MW, less likely to be used at such industrial settings, provided about 6,800 GWh (23,000 billion Btus) of thermal energy. And multiple market studies have concluded that a significant expansion of CHP system use within the building sector is possible--with over 35 GW of power generation potential in commercial buildings by 2020 (RDC 2005; EEA 2003a, 2003b). These studies note that current CHP systems in the commercial sector are concentrated in education and health-care buildings but that significant potential exists for many other building segments, including hotels. Hotels demand large amounts of coincident electrical and thermal energy on a year-round basis and are therefore likely to be a good match to a CHP system. Hotels also represent an opportunity for replicability because many units are owned by large corporations.

The CHP system described in this paper includes four microturbines producing a total of 227 kW of net rated capacity and a double-effect absorption chiller rated at 142 refrigeration tons (RT) (Wagner et al. 2007). Using a microturbine in a CHP system offers the well-documented advantages of fuller utilization of the available energy. For any simple Brayton power cycle, the decrease in cycle efficiency that occurs as the ambient temperature increases is well understood. This can be a detriment during hot summer months when the demand for electricity is often highest. However, it is not as well recognized that CHP systems based on a Brayton power cycle that include thermally activated equipment, such as the absorption chiller used in this project, actually have an increase in the overall system efficiency with increasing ambient temperature. This trend is due to a compensating increase in the efficiency of the CHP's thermally activated component caused by the increase in the turbine exhaust temperature. This has been documented in the laboratory (Zaltash et al. 2006) and is also shown in the field data reported here.

An important aspect of the system described in this paper is that the CHP system has been truly integrated by the manufacturer to match the capacities and performance parameters of each component. Previous research has shown the important relationships between overall system performance and the design and sizing of each component (Zaltash et al. 2005). These components include the microturbine, the absorption chiller, multiple heat exchangers, fuel gas boosters, and the control system. For example, the double-effect absorption chiller used here was originally designed to accept heat from a combustion chamber, but it was redesigned with an advanced heat exchanger to accept heat from the microturbine exhaust (Rosfjord et al. 2003). The double-effect absorption chiller also represents a careful balance between the increase in efficiency that occurs when moving from single to double to triple effect units and the corresponding increase in cost and complexity (Mone et al. 2001; Ziegler 2002).


The objective of the test and verification project described in this paper was to install, operate, and monitor a packaged CHP system at a deluxe hotel in San Francisco, CA. Information collected during the project was used to characterize the technical and economic performance of the CHP system under typical hotel operating conditions and describe lessons learned for use in future designs and installations.


The equipment manufacturer teamed with the research organization and the parent company of the hotel to build, deliver, install, test and demonstrate a microturbine-based CHP system at the hotel in San Francisco. The parent organization of the hotel strongly desired to employ this solution to offset its energy loads due to the system's energy efficiency, low emissions, and potential for replicability at other locations. The pre-engineered, integrated system provides base-load electrical power and space conditioning using four micro-turbines and a double-effect absorption chiller. Prior research established the technology for this system, including the diverter valve and control technology to independently meet electrical and cooling needs (Wagner 2004). The system performance was previously documented at specified operating conditions at the manufacturer's CHP laboratory, which contains load banks to simulate a wide range of thermal loads in accordance with established industry methodology (Rosfjord et al. 2003). Under the project described in this paper, the CHP system design was tailored to meet specific site requirements. Site integration issues were addressed in conjunction with the hotel, utility, and city. The CHP system was installed and monitored for a period of one year using a remote monitoring system.

CHP System Design

The CHP system contains four microturbines, each rated at 60 kW of electrical power at a 59[degrees]F (15[degrees]C) sea-level condition. Rated NOx emissions are less than 9 ppm at 15% exhaust oxygen, which met local emission requirements in force at the time of the installation. The exhaust from each microturbine is manifolded together to deliver input energy to a double-effect absorption chiller (Figure 1). The lithium bromide chiller consists of an evaporator, an absorber, a condenser, high-temperature and low-temperature generators, solution heat exchangers, refrigerant and solution pumps, a purge, controls, and auxiliaries. The chiller is an adaptation of a direct-fired chiller that increases the heat transfer area of the first-stage generator to compensate for the lower temperature inlet energy. Because it is a double-effect device, the chiller effectively converts the input thermal energy to chilled water and achieves a coefficient of performance (COP) of approximately 1.3. The double-effect feature also permits a manual changeover of the chiller to operate as either a chiller or heater. In this manner, the CHP system can become a "four seasons" product that provides either space chilling or space heating. The control system includes a diverter valve in the duct between the microturbines and the chiller. If the chilling demand is zero, this valve diverts the microturbine exhaust to atmosphere. If a chilling demand exists, the diverter is positioned to deliver the energy required for the chiller to meet the demand. The ability to isolate the chiller under no-load situations is important to avoid excessive concentrations within the chiller and possible solution crystallization. The diverter valve is designed-and was tested-to ensure this capability.


Also shown in Figure 1 are the fuel gas boosters (FGBs) that elevate the pressure of the natural gas fuel supplied by the gas utility to the level required by the microturbine. Each CHP system uses one FGB for a pair of microturbines. The FGB is powered by the direct-current power produced within one of the microturbine pair; therefore, that microturbine experiences a parasitic electrical load that reduces its alternating-current output. The other microturbine of the pair does not have this output reduction.

Table 1 details the performance specifications of the CHP system at 95[degrees]F (35[degrees]C) and at 59[degrees]F (15[degrees]C). The net power levels include power for the two FGBs. As indicated, the combined electrical and chilling capability of this project's CHP system results in CHP efficiency greater than 80%. To achieve this level in an application, the full system output capacity must by used productively by the building.
Table 1. CHP System Performance Specifications

                          Rated Performance

                   at 95[degrees]F  at 59[degrees]F
                   (35[degrees]C)   (15[degrees]C)

Net Power, kW           193              227
Cooling, RT             124              142
CHP Efficiency, %        80               91

Project Site and Chiller Integration

The CHP system was installed at a deluxe hotel in downtown San Francisco. The hotel is owned by a real estate investment trust whose portfolio includes over 100 properties in 26 states, including large holdings in California and Florida and near the cities of Atlanta, Boston, New York, and Washington, DC. The hotel recently completed a $12.5 million renovation and features 336 guest rooms. The owner desired to add a CHP system to the energy infrastructure because of its alignment with a corporate vision for environmental stewardship and the potential for energy cost savings.

Based on historical data and analyses, the hotel energy demand averages 670 kW of electrical power and 1200 kW of combined thermal energy use and power. Figure 2 traces instantaneous electrical power demand for a full year beginning in November 2002. The electrical demand during the year rarely dropped below 500 kW.


Because of the hotel's significant and persistent air-conditioning demand throughout the year, the CHP system was integrated only with the chilled-water loop (Figures 3a, 3b, and 3c). The absorption chiller operated in parallel with two existing 300 RT electric chillers (a primary unit and a spare). However, the design chilled-water flow rate for the building chilled-water supply loop was much higher than for the absorption chiller. To accommodate the different flow rates and pressure drops, a bypass loop with motorized isolation valves was required to balance flow rates during different operating modes.


In the absorption chiller only mode with CHP (Figure 3a), the motorized valves were positioned to allow returning chilled water to flow only through the absorber and the bypass loop. The chilled-water flow rate setpoint through the absorber was 270 gpm (17 L/s), measured by a flowmeter at the absorber exit. The bypass loop had a flow rate that varied according to the building chilled-water demand.

In the simultaneous chiller mode (Figure 3b), valve settings allowed flow through the absorption chiller and one of the electric chillers but not through the bypass. When this occurred, the lower flow resistance of the electric chiller reduced the chilled-water flow through the absorber to 170 gpm (11 L/s).

In the electric chiller only mode (Figure 3c), the valve positions isolated the absorption chiller and bypass loop. The chilled-water flow rate through the active electric chiller was roughly 500 gpm (32 L/s).

Installation in an Urban Environment

The CHP system was installed on the roof four stories above grade (Figure 4). The walled-in space was directly above the hotel mechanical room but was long and narrow and adjacent to an outdoor courtyard. The proximity to the mechanical room minimized plumbing integration challenges with the chilled-water loop, but the closeness to the courtyard required that the system be very quiet. Additionally, the ground-level space adjacent to the hotel available for hoisting the equipment was a narrow alley with private access. Each of these issues was resolved.


The CHP system ducting and FGB locations were conformed to the available rooftop space by modifying the standard package locations as necessary. All components were then positioned to leave the existing backup generator and cooling tower in their original locations.

Limited hoist access posed a significant challenge to getting the chiller onto the rooftop location. While each micro-turbine weighs only 1700 lb (770 kg), the absorption chiller weighed 18,500 lb (8400 kg). An initial crane hoist lifted the microturbines and FGBs without special equipment; a second lift for the chiller (Figure 5) was delayed a month to establish crane space requirements, city permits, and access to the private alley adjacent to the hotel for the larger crane.


Grid Interconnection

Most of the electrical grid in the US provides electricity to an end user using a radial connection. While there may be many branches in the cable from the power plant, the end user has one final radial feeder at the site. When a distributed generation source is located at such an end user, a reverse power relay is included in the radial interconnect to disconnect this on-site source to prevent electrical energy from being exported back onto the grid.

The electricity supply to end users in some metropolitan areas uses a network of multiple feeders to the sites. The multiple supplies provide redundancy in the electricity supply, enhancing power reliability. However, they also require network protectors on each utility feeder on the customer side of the transformer. A network protector is a combination of a breaker and a reverse-current protection relay to prevent the reverse flow of current onto a feeder that experiences a fault. Its purpose is to prevent the flow of electrical energy from one feeder back onto another feeder. The protectors are set to instantaneously detect the reversal and open the contactor, but that opening takes 5-25 seconds and requires a manual reset.

The hotel has a network electric utility connection and an on-site generator. This situation required special studies and electrical upgrades, and consequently, additional cost, to achieve a grid interconnection for the CHP system that was safe and reliable.

When on-site power generation is installed at a site with a network supply, it may be possible for the site load to momentarily drop below the generator output, resulting in an export of electricity unless other preventive devices are used. This possibility is minimized by requiring a buffer between the generator and the normal load. However, this measure does not guarantee that an export will never occur. If an export does occur, the network protector senses a reverse current and instantaneously begins to open. The utility company expressed concern that all network detectors might sense the reversal and begin to open, rendering the site without any grid-supplied electrical power and requiring time and cost to reset them. To avoid this situation, significant interconnection upgrades were required at this site (Figure 6).


The network protectors were upgraded with an adjustable time delay to avoid the instantaneous response and an under-power relay that opens if the net demand for grid electricity drops below a threshold of 25 kW. Additionally, a controller isolates the on-site generator if it senses that any one network protector has opened--either because of a feeder fault or power export--adding redundant protection to prevent reverse electricity flow to any feeder. The cost of the interconnect upgrades required by the utility totaled approximately $140,000. Implementation of a uniform interconnection standard has the potential to reduce this element of the project cost.

Data Acquisition System

The performance of the CHP system was measured for a period of one year using the manufacturer's remote monitoring system. The remote monitoring system provides for local and remote monitoring and control from either Intranet or Internet access. System alarms are automatically forwarded to a remote monitoring center for immediate response. Performance data and alarm history are stored in a central database. An internet web portal at the remote monitoring center provides connection to the site and allows for database queries and reports.

Data were acquired from remote monitoring system measurements made on each microturbine and the chiller and from additional instrumentation for this installation. Important parameters from each microturbine included the net electrical power and ambient temperature. The additional instrumentation measured the total flow rate of the natural gas fuel and the flow rate, return temperature, and leaving temperature of the chilled water at the absorption chiller.

The remote monitoring system acquired data every six minutes. Each entry consisted of three data records-one for the chiller, one for the four microturbines, and one for efficiency monitoring. Table 2 provides a subset of the recorded or calculated performance parameters for each of these data records. Each record had a date/time stamp. When all aspects of the remote monitoring system were operating properly, the three records had the same day/time stamp. However, there were instances of communication errors or other indicators of unreliable data in one or more of the records. All data records were reviewed to identify data gaps and to assure that data records were synchronized to the same date/time stamp when multiple records were required to assess a system feature. For example, electrical power was contained in the microturbine record. Therefore, assessments of microturbine output, runtime, and total delivered electrical energy required only the microturbine record. Chilled-water flow rate and return and leaving temperatures were contained in the efficiency monitoring record, allowing determination of the chiller output from this record alone. However, power generation and CHP efficiency calculations required both the microturbine and efficiency monitoring records to combine electrical, cooling, and fuel flow rate data.
Table 2. Measured and Calculated Performance Parameters zin
Remote Monitoring System

                          Chiller Record


   Run mode-always chilling                  Chilling or heating

         Run status                         Off, standby, running

Chilled-water return temperature                [degrees]F

Chilled-water leaving temperature               [degrees]F

       Microturbine Record (Items for Each Microturbine)


          Run time                                  h

Intel air temperature at compressor             [degrees]F

    Turbine exit temperature                    [degrees]F

       Engine shaft speed                          rpm

       Net output power                             kW

                Efficiently Moaltoring Record

     Fuel energy from flowmeter                   MBtu/h

Sum of microturbine net output power               kW

      Chilled-water flow rate                      gpm

 Chilled-water return temperature               [degrees]F

 Chilled-water leaving temperature              [degrees]F

     Calculated chilling rate                     MBtu/h

 Calculated electrical efficiency                   %

Calculated CHP efficiency                           %

Calculated cumlative fuel energy                  MMBtu

Calculated cumulative electrical energy            Mwh

Calculate cumulative chilling energy              MMBtu

Ambient temperature, calculated from            [degrees]F

microturbine inlet air temp at compressor

The energy efficiency of the CHP system was determined from the total energy delivered by the system and the fuel energy consumed to operate it. The electrical and chilling energy were determined by piecewise integrating the outputs presented previously over each time step for which valid data existed. The fuel energy was similarly determined from the integration of the measured natural gas flow rate and applying a constant heating value of 891 Btu/SCF. However, efficiency monitoring data were not reliably acquired until late May. Hence, for the first five months, the fuel energy rate was conservatively estimated as the quotient of the measured net electrical power and a constant value of electrical efficiency specified at 24.6%, the average of the instantaneous determinations during June.


Operating Hours

Overall, the CHP system achieved an extremely high level of availability with minimal outages. It produced at least 60 kW of net electrical power for 8231 hours, or for 94% of the year. Table 3 presents the monthly breakdown of operating, non-operating, and data gap hours. For the year, data gaps represented 2.8% of the available run hours. A sequence greater than 1 h was considered a "data gap" and the system was considered non-operational even though it may have actually been operating. Additionally, the system was documented not to be producing power for another 3.2% of the available hours.
Table 3. Monthly CHP System Operatin Profile

        Max   Operating     Non-operating   Data Gap

         h     h     %        H     %       h    %

Jan     744   718   96.5%    26   3.5%      0   0.0%
Feb     672   633   94.3%    25   3.8%     11   1.7%
Mar     744   601   80.8%   113  15.2%     30   4.0%
Apr     720   510   70.8%   109  15.1%    101  14.1%
May     744   648   87.0%     0   0.0%     92  12.3%
Jun     720   717   99.6%     0   0.0%      3   0.4%
Jul     744   742   99.7%     0   0.0%      1   0.1%
Aug     744   744  100.0%     0   0.0%      0   0.0%
Sep     720   717   99.6%     3   0.4%      0   0.0%
Oct     744   741   99.6%     0   0.0%      3   0.4%
Nov     720   718   99.7%     2   0.2%      0   0.0%
Dec     744   742   99.8%     0   0.0%      2   0.2%

Total  8760  8231   94.0%   277   3.2%    243   2.8%

There were four discernible periods when the microtur-bines were not producing power--approximately 20 h each on January 20, February 22, and March 20, and a 200 h period beginning March 28. Together, these four periods represent 94% of the non-operating time for the year. The first three were minor maintenance periods, including instances when the system was turned off to investigate chiller issues or system integration with the hotel chilling loop. The longest period was to repair the diverter valve in late March.

Most data gaps occurred during late April and were due to communication interruptions between the site and the data server. Most of these occurred for a full day; one lasted slightly longer. No significant data gaps occurred in the second half of the year.

In contrast with the first half of the year, periods of non-operation and data gaps were almost non-existent during the second half of the year. Electrical energy was delivered to the hotel for over 99.7% of total hours in these months.

Delivered Power and Chilling

Figures 7 and 8 depict the instantaneous net electrical power and chilling delivered to the hotel for the first and second halves of 2006, respectively. The net electrical power includes the reduction due to the parasitic load of the two FGBs. Aside from the outage to repair the diverter valve in late March, the microturbines delivered full power throughout the first half of the year. Three periods of half power occurred during the second half of the year--starting July 4, September 8, and November 13. In each case, the reduced power was due to an FGB failure, which took two microturbines offline each time.


Root cause analyses indicated that "black powder" in the natural gas, a mix of fine particulates including ferrous material, shorted the FGB motor. A motor redesign to isolate the motor from the natural gas was considered but was not implemented during the evaluation period. An alternative solution used a higher-temperature rated motor. The associated enhanced wire insulation appears to have successfully prevented further FGB failures.


Net power output responded as expected to daily and seasonal ambient temperature variations. This behavior reflected the expected response of turbine engine power derate as ambient temperature increased. Microturbines #2 and #3 did not power an FGB. At lower ambient temperatures they delivered a controlled output of 60 kW. Output fell off as expected beginning near 70[degrees]F (21[degrees]C) at a rate of 1 kW for every 3[degrees]F (1.7[degrees]C) rise in ambient temperature. The measured output at 95[degrees]F (35[degrees]C) was slightly higher than rated output. Microturbines #1 and #4 each powered an FGB. Their net power was reduced by approximately 10% because of the FGB electricity requirement.

The absorption chiller delivered at least 10 RT to the hotel for 7557 hours in 2006, 92% of the microturbine operating hours. Four main periods of unscheduled chiller outage (in April, July, October, and November) totaled 390 h during microturbine operation. The first of these was associated with diverter valve replacement and the last two were for diverter valve adjustment. The other was associated with the July FGB failure.

Delivered chilling varied greatly, both daily and seasonally. These fluctuations reflected the changing chilling demand by the hotel, which was set by a combination of hotel activity and ambient temperature. While the CHP system had the capability to deliver over 120 RT, the actual chilling output was set to the level demanded by the hotel from the CHP system. This level nominally ranged between 50 and 125 RT, which was achieved, for nearly constant full-power electrical output, by modulating the diverter valve to deliver only the input energy required to match the chilling demand.

Figure 9 shows delivered power and chilling relative to ambient temperature during two typical winter days in January. Cooler ambient temperatures during the mild winter in San Francisco caused only a minor variation in the net delivered power. For the ambient temperature of 59[degrees]F [+ or -] 5[degrees]F (15[degrees]C [+ or -] 2[degrees]C), delivered power was stable at 239[+ or -]3 kW. However, delivered chilling ranged from 45 RT at night to 85 RT during the afternoon. For these ambient temperature levels, the average hotel chilling demand was only 45% of the system chilling capacity.


Figure 10 illustrates generation efficiency (multiplied by 10 on this graph), delivered power, and chilling relative to ambient temperature during two typical summer days in June. As expected, warmer ambient temperatures of 70[degrees]F [+ or -] 8[degrees]F (21[degrees]C [+ or -] 4[degrees]C) caused a 20 kW reduction in electrical power each afternoon when the ambient temperature reached its peak. Generation efficiency decreased by approximately 1% at higher afternoon temperatures. Delivered chilling ranged from 62 RT at night to 125 RT during the afternoon. Peak chilling demand was 93% of system capacity.


The difference between January and June chilling demands was driven by only an 11[degrees]F (6[degrees]C) increase in average daily temperature and a 14[degrees]F (8[degrees]C) increase in peak temperature. The average chilling demand increased 55%, indicating its strong dependence on ambient temperature as well as internal loads.

In addition to hotel demand, the chiller output was limited by two other factors. One limitation occurred during periods when two of the microturbines were not operating because of a FGB failure. The reduced microturbine exhaust energy limited the maximum chilling output to less than 75 RT. The other limitation was associated with the integration of the CHP system with the pre-existing on-site electrical chiller. The control strategy implemented for this system resulted in repeated shifts between the CHP mode (Figure 3a) and the simultaneous mode throughout the year.

The switch from CHP mode to simultaneous mode occurred whenever the absorber output alone could not satisfy the hotel demand for chilling. In this case, the absorber capacity could not suppress the chilled-water temperature returning from the hotel to the desired setpoint for the chilled-water temperature required to cool the hotel. The chilled-water temperature leaving the absorber is typically used for system control when a parallel chiller is not present. However, for parallel operation with the electric chiller, the returning temperature was a better indicator that the absorber was not keeping up with the demand and that simultaneous mode should be initiated.

The general sequence when switching from CHP mode to simultaneous mode was:

1. The absorption chiller output satisfied the hotel demand as indicated by stable and acceptably low returning temperature.

2. As the hotel demand grew, the diverter valve closed to deliver increasing energy to the absorber and to try to maintain absorber supply temperature. As the load continued to grow, both return and supply temperatures increased.

3. When the returning temperature exceeded a "high" setpoint, the motorized valves activated and the electric chiller started to achieve the simultaneous mode.

4. The absorber chilled-water flow rate dropped from 270 to 170 gpm (17 to 11 L/s) due to the lower flow resistance of the electric chiller. The lower demand on the absorber required the diverter valve to open to bypass in order to maintain the absorber supply temperature setpoint even though the return temperature was high.

5. The hotel demand was not satisfied until the electric chiller output and the reduced absorber output stabilized the return temperature.

6. CHP mode was re-established after hotel demand reduced sufficiently to allow the return temperature to drop below a "low" setpoint (5[degrees]F [3[degrees]C] lower than the "high" setpoint to reduce mode cycling).

When switching from CHP mode to simultaneous mode, the lower chilled-water flow rate through the absorber reduced its chilling output. In addition, when the electric chiller started, it did not "shave" the chilling peak demand. Instead, it suppressed the absorber output to below what it had been delivering, reducing the use of "free" CHP chilling and reducing the overall CHP efficiency. The absorber output suppression occurred because the electric chiller did not have the ability to continuously turn down to 0 RT but had a minimum capacity of approximately 75 RT. The total chilling stepped from 115 RT for the absorber alone to 155 RT (80 + 75 RT) with both chillers operating. This was sufficient output to begin to suppress the returning chilled-water temperature and subsequently reduce the absorber output even further.

Figure 11 illustrates this effect during July 28-29, when all four microturbines were operating and the system switched between CHP mode and simultaneous mode. The electrical power was near 220 kW, reducing slightly during the afternoon of July 28 when a peak ambient temperature of 75[degrees]F (24[degrees]C) was experienced. The returning temperature rose to 57[degrees]F (14[degrees]C) as the ambient temperature increased, causing the system to switch to simultaneous mode. The chilled-water flow rate through the absorber then dropped from 265 to 165 gpm (17 to 11 L/s). The leaving temperature, which was also rising in CHP mode, dropped when the mode switched and was maintained at 43[degrees]F (6[degrees]C). The chiller output was 115 RT before the mode switch then dropped to 80 RT because of the reduced chilled-water flow rate.


Prior to switching to simultaneous mode, the combined outputs resulted in a maximum CHP efficiency of 72%, which immediately dropped to 55% after the switch. Simultaneous mode was sustained for 12 h. During this time, the return temperature continued to decrease because of the electric chiller output and the decreasing ambient temperature. With the simultaneous mode control logic and the fixed minimum electrical chilling capacity, the absorber chilling output varied to match the decreasing hotel load, which also decreased the CHP efficiency.

When the chilled-water return temperature dropped below 52[degrees]F (11[degrees]C), the system returned to CHP mode, shutting off the electric chiller. The chilled-water flow rate to the absorber chiller increased back to 265 gpm (17 L/s) and the absorber output rose to meet the hotel demand. The chilled-water supply and return temperatures increased in response to the rising ambient temperature the next day. The switch to simultaneous mode was repeated starting in late morning of July 29, with similar impacts.

The average delivered power during the year, determined by dividing the total electrical energy by the total microturbine operating hours, was 211 kW. The average chilling output was 74 RT. While both values were slightly reduced because of the periods when only two microturbines were operating, the electrical power was within 7% of the system rating, affirming that full power output was typically demanded by the hotel and delivered by the system. In contrast, the average chilling output was only 52% of the system design capacity because of the lower average demand on the CHP system and its strong day-to-night variation.

CHP Efficiency and Operating Economics

Table 4 displays the values of delivered and consumed energy for each month of 2006 as well as the calculated level of electrical and CHP efficiency (defined as the sum of electrical and chilling energy divided by the fuel energy). As described previously, the estimated electrical efficiency for January through May was lower than for any other period and hence slightly conservative. The CHP efficiency was highest during the third quarter because of the greatest demand for chilling.
Table 4. Monthly CHP System Energy Efficiency Profile

              Energy, MMBtu               Eficiency, %

       Fuel    Electrical  Chilling   Electrical     CHP

Jan    2,371      582         586       24.6%       49.3%
Feb    2,061      506         566       24.6%       52.0%
Mar    1,966      483         326       24.6%       41.1%
Apr    1,628      400         322       24.6%       44.3%
May    2,011      494         562       24.6%       52.5%
Jun    2,071      531         750       25.7%       61.9%
Jul    1,703      430         613       25.2%       61.2%
Aug    2,152      546         611       25.4%       53.8%
Sep    1,307      350         608       26.8%       73.3%
Oct    2,145      550         677       25.9%       50.1%
Nov    2,004      519         486       25.9%       50.1%
Dec    2,129      547         597       25.7%       53.7%
Total  23,547   5,938       6,703       25.2%       53.7%

Table 5 presents the 2006 operating economics for the CHP system based on delivered energy and utility bills. The electricity offset by chilling is based on an assumed electric chiller COP of 4.0. A constant monthly maintenance cost was also included. For the year, the CHP system reduced energy costs by $73,560. Electric rate increases that occurred after the demonstration time period have increased the savings to about $120,000/year. The total installed cost of the system was $1,040,000. Based on the experience gained during this project, the equipment manufacturer projects the installed cost for the next similar project will be about 10% less and will eventually drop by about 25% as more installation experience accumulates (Morrow 2006).
Table 5. Monthly CHP System Operating Economics Profile

         Electrical     Chilling    Electricity    Avioded
          Energy         Energy      Offset by     Utility
         Delivered     Delivered     Chilling    Electricity

        MMBtu   MWh   MMBtu   MWh      MWh          MWh

Jan      582    171    586    172       43          213
Feb      506    148    566    166       41          190
Mar      483    141    326     95       24          165
Apr      400    117    322     94       24          141
May      494    145    562    165       41          186
Jun      531    156    750    220       55          211
Aug      546    160    611    179       45          205
Sep      350    103    608    178       45          147
Oct      550    161    677    198       50          211
Nov      519    152    486    142       36          188
Dec      547    160    597    175       44          204

Total  5,937  1,740  6,703  1,964       491       2,231

       Electricity   Fuel Cost
         Savings        and         Energy
                    Maintenance     Savings

           $             $            $

Jan     $29,886      $(26,296)      $3,589
Feb     $26,556      $(23,239)      $3,317
Mar     $23,136      $(17,978)      $5,158
Apr     $19,696      $(17,831)      $1,865
May     $26,020      $(20,903)      $5,117
Jun     $29,484      $(20,166)      $9,318
Aug     $28,672      $(16,869)      $7,034
Sep     $20,597      $(14,123)      $6,474
Oct     $29,516      $(17,777)     $11,739
Nov     $26,269      $(21,952)      $4,317
Dec     $28,545      $(20,732)      $7,813

Total  $312,280     $(238,720)     $73,560


The CHP system operated for 8231 hours during the one year monitoring period, or 94% of the time. Ninety percent of these hours were at full power. A careful analysis of two fuel gas booster failures that occurred during the first year of operation resulted in the specification of more durable motors within these system elements. The electrical energy capability of the CHP system was fully demanded by the hotel. Instantaneous CHP efficiency often exceeded 70%. However, the hotel chilling demand was limited by the low demand during cooler months and nighttime hours and by the interaction with a parallel, pre-existing electric chiller. This interaction suppressed CHP chilling because of a high minimum electric chiller output. For brief periods when only two microturbines operated, the higher proportional use of the available CHP chilling resulted in a CHP efficiency averaging over 80%, with frequent instances exceeding 90%. Over the entire test period, the reduced use of the available thermal energy limited the CHP output to 52% of its capability. Within this limitation, the system delivered 1.74 GWh of electrical energy and 6700 MMBtu (1.96 GWh) of chilling energy while consuming 235,000 therms of natural gas fuel energy. For the year, the CHP efficiency was 53.7%. Changes were made after the test period to improve integration with the building and increase energy efficiency.

The hotel owner and operator were pleased with the CHP system, as evidenced by their orders for two more systems. They were pleased with the reliability of operation, the energy cost savings achieved, and the environmental pollution avoided. While not monetized, the latter was consistent with the owner's commitment to environmental stewardship. Due to the lower-than-anticipated thermal usage, the annual savings during the demonstration period were only $74,000. Electric rate hikes since that time increased the savings to $120,000/year. Based on lessons learned and recent product enhancements, a future installation under similar conditions would be configured to provide simultaneous heating and cooling rather than cooling alone; this would increase the thermal utilization and provide additional savings by offsetting expensive steam heating. The projected savings for such a system would be $250,000/year, resulting in a 4.2 year payback without incentives.


The project described in this paper was developed within the National Accounts Energy Alliance in response to a solicitation issued by the Oak Ridge National Laboratory on behalf of the U.S. Department of Energy. Under the resulting collaboration between the Department of Energy, Gas Technology Institute, and Oak Ridge National Laboratory, UTC Power with Host Hotels and Resorts to install and operate a combined heat and power system at the Ritz-Carlton, San Francisco.


ConEdison. 2007. A brief history of Consolidated Edison. New York: ConEdison.

EEA. 2003a. Market potential for advanced thermally activated BCHP in five national account sectors. Energy and Environmental Analysis, Inc., Arlington, VA.

EEA. 2003b. Task 2.1 report: National account sector energy profiles. Energy and Environmental Analysis, Inc., Arlington, VA.

Hadley, S.W., J.W. Van Dyke, and T.K. Stovall. 2003. The effect of distributed energy resource competition with central generation. ORNL/TM-2003/236, Oak Ridge National Laboratory, Oak Ridge, TN.

Mone, C.D., D.S. Chau, and P.E. Phelan. 2001. Economic feasibility of combined heat and power and absorption refrigeration with commercially available gas turbines. Energy Conversion & Management 42:1559-73.

Morrow, A., and P. Savarino. UTC Power PureComfort 240 Ritz-Carlton San Francisco. Oak Ridge National Laboratory, Oak Ridge, TN.

Petrov, A., et al. 2006. Commercial integrated energy systems provide data that advance combined cooling, heating, and power. IMECE2006-14932. Proceedings of the 2006 ASME International Mechanical Engineering Congress and Exposition.

RDC. 2002. Integrated energy systems (IES) for buildings: A market assessment. ORNL/SUB/409200. Resource Dynamics Corporation, Vienna, VA.

RDC. 2005. DG Monitor: The Installed Base of U.S. Distributed Generation. Vienna, VA: Resource Dynamics Corporation.

Rosfjord, T., et al. 2003. Research, development, and demonstration of packaged cooling, heating, and power systems for buildings. DOE/DER Peer Review, December 2003,

DOE. 2007. Buildings Energy Data Book. Washington, DC: U.S. Department of Energy Office of Energy Efficiency and Renewable Energy.

Wagner, T. 2004. Energy-savings systems for commercial building CHP and industrial waste heat applications. Cogeneration and Distributed Generation Journal 19(4).

Wagner, T., T. Rosfjord, and A. Morrow. 2007. National account energy alliance final report for the field scale test and verification of a PureComfort[R] 240M combined heat and power system at the Ritz Carlton, San Francisco. ORNL/TM-2007/101, Oak Ridge National Laboratory, Oak Ridge, TN.

Zaltash, A., et al. 2005. Integration of an indirect-fired absorption chiller in the microturbine-based CHP System. ISHPC-092-2005. Proceedings of the International Sorption Heat Pump Conference.

Zaltash, A., et al. 2006. Laboratory R&D on integrated energy systems (IES). Applied Thermal Engineering 26(1):28-35.

Ziegler, F. 2002. State of the art in sorption heat pumping and cooling technologies. International Journal of Refrigeration 25(4):450-59.

Timothy C. Wagner, PhD


Neil P. Leslie, PE


Richard S. Sweetser


Therese K. Stovall, PE


Timothy C. Wagner is a principal engineer at United Technologies Research Center, East Hartford, CT. Richard S. Sweetser is president of Exergy Partners Corp., Herndon, VA. Neil P. Leslie is a research manager in the End Use Solutions Sector at Gas Technology Institute, Des Plaines, IL. Therese K. Stovall is a senior research engineer in the Engineering Science and Technology Division at Oak Ridge National Laboratory, Oak Ridge, TN.
COPYRIGHT 2009 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Wagner, Timothy C.; Leslie, Neil P.; Sweetser, Richard S.; Stovall, Therese K.
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2009
Previous Article:Laboratory measurements to quantify the effect of bypass on filtration efficiency.
Next Article:Recovered energy generation using an organic Rankine cycle system.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters