Designing Hydronic Systems to Maximize Efficiency and Response.
Overall hydronic heating system efficiency will be affected by the efficiency and interactions of each component chosen. While the heating appliance may come with an efficiency rating as high as 95%, when pump energy, pipe and boiler jacket losses, and boiler settings are factored in, the overall system efficiency could be significantly less. Based on previous research efforts, it is apparent that hydronic systems are typically not designed and installed to achieve maximum efficiency, especially when paired with baseboard convectors (Arena 2010): one of the most cost-effective emitters for hydronic systems. This research effort was intended to identify combinations of components and controls that result in the best overall system efficiencies.
Equipment specifications along with the method for supplying domestic hot water (DHW) and the minimum and maximum space heating output capacities are displayed in Table 1.
System 1 consists of a high mass, condensing water heater which provides both space heating and DHW. The water in the tank is supplied directly to the DHW end uses. When there is a call for heat by the thermostat, space heating is provided by cycling the water in the tank through a brazed plate heat exchanger which transfers heat to the heating loop. This heated water is then circulated to the conditioned space and distributed via baseboard convectors. This appliance has a 5 to 1 turndown ratio on the main combustion system and a 10 to 1 turndown ratio for space heating resulting in a minimum space heating output capacity of 10,000 Btu/h (2.9 kW). The additional reduction in space heating capacity is achieved through modulation of the pump feeding the brazed plate heat exchanger.
The outdoor reset curve for System #1 was initially programmed to operate as indicated in Figure 1. At an outside temperature of 0[degrees]F (-18[degrees]C), the space heating supply temperature would reach its maximum setting of 140[degrees]F (60[degrees]C) with a tank temperature setting of 145[degrees]F (63[degrees]C). The minimum water temperature supplied by the space heating module was set to 100[degrees]F (38[degrees]C.)
The variable speed circulator within the boiler monitors the supply water temperature to the space heating distribution system, and attempts to maintain the target supply water temperature as dictated by the outdoor reset curve. A second variable speed pump circulates the heated water to the zones. This pump was set to a constant pressure setting resulting in a flow rate of 1.0 gpm (0.06 L/s) to each zone.
System #2 is a 95 AFUE modulating, condensing, low-mass boiler that provides the DHW via an indirect tank. This system was set up to compare two different approaches to providing hydraulic separation: a buffer tank and a standard primary/secondary plumbing configuration. Ball valves were installed to switch the flow between the two plumbing configurations, and a toggle switch was used to direct the boiler's controller to either the temperature sensor in the buffer tank or one in the primary loop. The outdoor reset control was programmed to deliver 140[degrees]F (60[degrees]C) water to the baseboard at 0[degrees]F (-18[degrees]C) outdoors and 95[degrees]F (35[degrees]C) supply water when the outdoor temperature was 68[degrees]F (20[degrees]C).
This boiler was selected because it is equipped with controls intended to prevent short cycling and decrease recovery time. The installer can limit the maximum space heat input of the boiler up to 50%, if desired. This feature is beneficial in circumstances where the boiler is oversized for the space heating load either because the DFfW demand is larger than the space heating load, or the space heating loads are simply smaller than the smallest available boiler on the market. Another feature available with this boiler is a boost control which can be set to override the outdoor reset control if the heating demand is not satisfied within a specified amount of time as determined by the installer.
A total of four pumps were used in this system: one to circulate hot water to the DHW tank, one to circulate hot water through the buffer tank or primary loop, and two to circulate hot water to the space heating zones. All zone circulators have electronically commutated motors (ECMs). These pumps operate under constant differential pressure allowing them to react to changes in system pressure by increasing or decreasing the flow. The pumps were intended to provide 1.5 gpm (0.09 L/s) through each heating zone loop and approximately 5 gpm (0.32 L/s) through the primary loop/buffer tank and the indirect DHW tank.
Both systems were set up to capture total energy input to the heating system (gas and electricity), total heat output to the conditioned space, output to each heating zone, and output for DHW generation. To calculate these values, gas input, water flow rates, system temperatures, electric energy use, and indoor and outdoor temperatures were recorded. All sensors were hard wired to a data logger. Data was recorded every minute and transmitted daily via wireless modem. Table 2 lists the parameters measured and the location of the sensors.
The intention of this research effort was to evaluate different high efficiency hydronic systems and their corresponding control and plumbing configurations to determine the most efficient combination of components, therefore, several combinations of control settings, thermostat settings and plumbing configurations were evaluated as follows:
* Constant space temperature + primary loop
* Constant space temperature + buffer tank
* Night time thermostat setback + primary loop w/ boost control
* Night time thermostat setback + buffer tank w/ boost control
Following each change, the system was allowed to operate for approximately two weeks. System #1 was maintained in constant temperature mode for the majority of the test period because it was unable to adequately recover from setback. The space heating input was capped at 80% of the maximum input for System #1 and 50% for System #2 for the entire monitoring period.
Note that the house with System #1 was unoccupied for 3 months of the 5 month monitoring period, and the house with System #2 was vacant the entire test period.
Although the combustion efficiency of each heating appliance is much higher, the overall system efficiencies shown in Table 3 indicate that over 30% of the total input to space heating and DHW is being lost resulting in system efficiencies in the mid to upper 60%'s. However, measurements taken with a combustion analyzer during system commissioning revealed combustion efficiencies in the mid to upper 90%'s. Even when the condensing water heater was operated above ideal conditions for condensing (>130[degrees]F/54[degrees]C), the combustion efficiency was in the high 80%'s.
Besides space heating losses to the basements, standby losses associated with DHW production were also substantial. Table 4 summarizes the energy required to replenish standby losses from the storage tanks in System #1 and System #2 and the associated costs. Periods with no hot water draws and no space heating activity were examined to isolate these values. Replenishment of standby losses occurred approximately 3 times a day for each house.
A comparison of the results for various control configurations tested in System #2 is presented in Table 5. There is little difference in the results between primary/secondary loop and buffer tank operation aside from the amount of cycling. Boiler efficiency, overall system efficiency and operating costs are all very similar. The difference between constant temperature operation and night time setback are more pronounced. Even though efficiencies are similar between the two operating modes, setback results in a 16% savings in gas input and operating costs over constant temperature operation.
Decreasing System Losses Through Proper Design
When evaluating the overall system efficiencies, losses to the unfinished, unconditioned basements were seen as losses, not useful heat, and were not included in the space heating output unless the basement was warmer than the 1st floor as often occurred in the house with System #2. Heat transfer to the 1st floor was then estimated and added to the space heating output. This point could easily be countered, but it was not the intention of this study to define useful heat. What was found was that system losses were primarily the result of three problems.
First, the baseboards installed were not performing as anticipated. It is suspected that the overall system efficiencies would have been higher if the installed baseboard capacity had been higher allowing for lower temperature operation. These systems were designed to meet the design heating load using 140[degrees]F (60[degrees]C) supply water at 1.5 gpm (0.09 L/s). But, because the output of the baseboard was lower than expected, supply temperatures had to be increased, thereby, increasing gas input and standby losses. For system #1, that meant keeping the water in the tank at a higher temperature than desired for condensing operation. Once the tank temperature exceeds 130[degrees]F (54[degrees]C), condensing is severely limited as was confirmed with the use of a combustion analyzer during a period when the tank was being maintained at 144[degrees]F (62[degrees]C). The resulting efficiency reading on the meter was 89%, not 96% as recorded when the tank was being maintained at 120[degrees]F (49[degrees]C). The decreased baseboard capacity also affected System #2 because the boiler's boost control overrode the outdoor reset control more often so the boiler could provide warmer water to the zones to compensate for the lack of capacity. Increased baseboard capacity would have resulted in 1) shorter system run time equating to less pump and boiler fan energy consumed, 2) lower temperature operation which in turn would have resulted in less heat loss through pipes and storage tanks and increased combustion efficiency, and 3) reduced short cycling because the temperature difference between the boiler's supply and return temperatures would have been larger preventing the burner from turning off as often.
Second, the standby losses from the exposed piping and storage tanks in the basements were significant and exacerbated by the fact that the systems had to be operated at higher temperatures than expected to meet the heating demand and to recover from setback. In fact, the hydronic system losses to the basement from System #2 were so substantial that calculations predict a positive heat flow from the basement to the conditioned space (Figure 2). Reduced run time of the heating loop to zone one and the comparison of total output (8.7 MMBtu/2.5 MWh) to the calculated heat loss (10.3 MMBtu/3.0 MWh) for the entire monitoring period, support the assumption that the heat loss from the system to the basement was at least, in part, being transferred to the first floor. Other than solar gain, there were few internal gains available to offset the load. These losses could be drastically reduced by insulation and lower temperature operation.
Lastly, both homes were unoccupied for most or all of the monitoring period. This means that any energy to maintain the DHW tank temperature has to be considered standby loss. When occupied, these losses will become a smaller percentage of the overall input resulting in slightly higher overall system efficiencies. If the DHW input and output are eliminated from the efficiency calculation for System #2, the overall system efficiency increases to approximately 71%.
High Mass vs. Low Mass Operation
Both systems had added mass. A 55 gallon condensing water heater was the heating appliance installed in System #1 and an additional 30 gallon buffer tank was installed between the boiler and the space heating zones in System #2. Because the system in System #1 already provides hydraulic separation, it didn't require the standard primary/secondary loop plumbing typically found in low mass systems. This means that one less pump and less piping were required as compared to a low mass boiler with an indirect tank. The buffer tank in System #2 resulted in added costs over the standard primary/secondary loop configuration. There was no reduction in number of pumps and piping needed, and the buffer tank introduced an additional cost.
Boiler cycling, pump energy use and system response time were evaluated along with overall system efficiency during each mode of operation. The differences in these values were not significant enough to conclude that one operated more efficiently or resulted in lower operating costs, however, the buffer tank definitely reduced cycling on the order of 50% compared to the primary/secondary loop configuration. The benefits of reduced cycling, however, are not well documented. There are claims that short cycling reduces boiler life and efficiency, but no conclusive evidence was found. There is also conflicting opinions about the effects. Two different DOE publications were found: one stating cycling had little effect on efficiency (Henderson et al, 1999), the other claiming a significant effect (EERE 2012). There are no conclusive results from this research that support the claims that cycling reduces efficiency.
The boiler in System #2 was also equipped with additional controls intended to reduce cycling when used in primary/secondary operation. The first, limiting the boiler's space heating input, is intended to prevent the boiler from going to high fire on start up, limiting the speed at which the supply temperature will reach the setpoint. The second control reduces or increases the boiler's response time. The slower the response, the longer it will take for the supply temperature to reach the setpoint. When the system was tested with these controls in place, buffer tank operation resulted in the least amount of cycling. To evaluate the special controls more closely, buffer tank operation was then run with the factory default settings and compared to both buffer tank and primary/secondary loop operation with a 50% limit on the space heat input and a very slow response time. Using the defaults resulted in a 50% increase in cycling during buffer tank operation. However, buffer tank operation still resulted in 50% less cycling when compared to primary/secondary loop configuration with the controls set.
High Efficiency Pumps
Each system used high efficiency pumps with ECMs. The resulting savings over a system with standard pumps is 30% to 50%. However, careful attention to system head loss and pump performance curves is mandatory if high efficiency pumps are incorporated into the system design. The systems designed here had much higher system pressures than anticipated. It is suspected that the specialized fittings for the temperature sensors in the second floor baseboard restricted the flow more than the normal fittings would have. The ECM pumps installed were not able to provide more than 1.5 gpm (0.09 L/s) to the second floor zones in either system when set to their maximum speed. To reduce the electric energy use of the pumps, flow rates were set to 1.0 gpm (0.06 L/s) instead of the design value of 1.5 gpm (0.09 L/s), further reducing the capacity of the system.
90%'s, long-term data show that standby losses to unconditioned spaces can be as large as 20% to 30%. These losses are primarily due to standby losses from the water storage tanks, exposed piping and boiler jackets. It is recommended that installers insulate all exposed supply and return piping to the space heating zones and to the indirect DHW tank, if applicable, especially in unconditioned spaces. Additional insulation is also recommended for storage tanks located in unconditioned spaces.
Systems operated with thermostat setback and boost controls for recovery operate more efficiently, cycle less and use less energy than systems operated in constant temperature mode. It must be noted that in order for a hydronic system to recover from setback, the system must have the extra capacity needed to recover. The systems should be designed to provide the necessary output under design conditions at supply temperatures lower than the boiler's maximum supply temperature. For example, if the boiler's maximum supply temperature is 180[degrees]F (82[degrees]C), the system should be designed to meet the design load at 150[degrees]F (66[degrees]C). Then, when recovery is needed, the boiler can boost to higher temperatures to meet the added load. Both the boiler and the baseboard need to be able to provide this added output. For modulating, condensing boilers, design the system to meet the design heating load such that the return temperature does not exceed 13.0[degrees]F (54[degrees]C).
It was clearly observed in this research that added mass resulted in less cycling of the boiler's burner if all conditions are held equal--burner input, boiler response rate, thermostat control, etc. However, no significant differences were found in efficiencies or operating costs between primary/secondary loop and buffer tank operation. Therefore, installing a buffer tank for this particular system does not appear to provide a quantifiable benefit. If the boiler were severely oversized, a buffer tank might provide a more obvious benefit.
When using a condensing water heater for space heating, maintaining a condensing water heater above 130[degrees]F (54[degrees]C) will result in decreased combustion efficiency, and should be avoided by designing the system for low-temperature operation. In general, increases in mass temperatures, as may be needed for space heat operation, will result in increased standby losses. Added tank insulation should be considered, especially for storage tanks located in unconditioned spaces. Design the system to meet the design heating load such that the tank temperature does not exceed 130[degrees]F (54[degrees]C).
Inadequate baseboard capacity will result in extended run times, increased standby losses and an overall decrease in system efficiency. Baseboard should be sized for low temperature operation under design conditions and be able to provide twice that capacity under boost operation to reduce recovery time from thermostat setback.
High efficiency pumps prove to be cost-effective enhancements and have a simple payback between 4 and 5 years. Pumps that have displays showing flow rate and energy consumption are particularly useful during commissioning to optimize performance and ensure design conditions are being met. Careful attention must be given to the pump performance curves and the maximum system head loss the pump can handle.
The authors would like to thank DOE and all the manufacturers and product suppliers who donated their time, expertise and equipment: Dave Davis from Heat Transfer Products, Dave Scearce from Peerless, Mark Handzel with Bell & Gossett, Bob Reimund from Grundfos, and Parker Wheat with Emerson Swan. A special thanks to Scott Reynolds with Ithaca Neighborhood Housing Services for his continued support and the use of his homes for this research.
Arena, L. 2010. "In-Field Performance of Condensing Boilers in Cold Climate Region." National Energy Technology Laboratory, Morgantown, WV; US Department of Energy, Building Technologies Program.
Butcher, T.; (2006). "Condensing boilers and baseboard hydronic systems.", ASHRAE Transactions, V. 112, part 1, 2006.
Butcher, T; (2009). "Optimal Design and Operating Parameters of the Condensing Boiler/Hot Water- Baseboard Combination: Task Report." NYSERDA Agreement, No. 10927. Albany, NY; NYSERDA.
Henderson, H., Huang, Y., Parker, P. 1999. "Residential Equipment Part Load Curves for Use in DOE-2". LBNL-42175. US Department of Energy, Office of Building Systems, Contract No. DE-AC03-76SF00098.
Lois B. Arena
Associate Member ASHRAE
Lois B. Arena is a Senior Engineer with Steven Winter Associates, Inc., Norwalk, CT. Omari Faakye is an EIT with Steven Winter Associates, Inc., Norwalk, CT.
Table 1. Mechanical System Specifications. Component System 1 98% Combustion Space Heating Efficiency Condensing Water Heater w/ baseboard convectors DHW 55 gallon (208 L) tank Space Heating Output 100,000 Btu/h (29 kW) (max/min) 10,000 Btu/h (3 kW) Design Heat Load 14,695 Btu/h (4 kW) Component System 2 95.1% AFUE Condensing Boiler Space Heating (buffer tank vs. primary/secondary loop) w/ baseboard convectors DHW Indirect, 37 gallon (140 L) tank Space Heating Output 46,000 Btu/h (13 kW) (max/min) 14,720 Btu/h (4.3 kW) Design Heat Load 15,426 Btu/h (5 kW) Table 2. Monitoring Setup Parameter Sensor Location Gas Input Main gas line into the boiler Pump Electrical Energy At each pump Boiler Electrical Energy At main electrical line into the boiler Boiler Supply Return Temps Inlet and outlet temperatures on boiler main supply and return piping Zone Supply/Return Temps Right before the zones split from the primary loop and on each return line before the primary loop System Flow Rates Each zone, the primarily loop and the cold water from the main into the water heater DHW temperatures On the hot water line right after the boiler/DHW tank and on the main cold water line into the system. Interior Temperatures Each bedroom, kitchen, living room, dining room and basement. Exterior Temperature At the outdoor reset sensor location. Table 3 Space Heating and DHW Input and Output Summary End Use Measured Value System #1 System #2 Space Heat Total Output to Zones (MMBtu/MWh) 9.9/2.9 8.7/2.6 DHW Total Output (MMBtu/ MWh) 0.8/0.2 0.2/0.06 Total System Gas Input (MMBtu/ MWh) 15.5/4.5 15.0/4.4 Entire System Elec. (kWh) 131.3 109.3 Total System Input (MMBtu/ MWh) 16.0/4.7 15.4/4.5 Total System Output (MMBtu/ MWh) 10.7/3.1 10.0/2.9 Overall System Efficiency 67% 65% Table 4 Energy Used to Replenish DHW Standby Losses End Use System #1 System #2 Energy Consumption per Replenishment Cycle (1) [Btu (kWh)/cycle] 2500 (0.7) 6000 (1.8) Annual Energy Consumption [MMBtu (MWh)/yr] 2.7(0.8) 6.6 (1.9) Annual Cost (2) $38 $92 (1) Measurements take for a tank temperature setpoint of 120[degrees]F in System #1 and 130[degrees]F in System #2. (2) Assumed $1.34/therm. Table 5 Comparison of Performance for Each System Configuration in System #2 All Values Normalized per HDD Variable Control Settings Primary Butter Tank Loop Space Heat Gas Input (Btu/kWh) 5,729/1.68 5,846/1.71 Space Heat Pumps (kWh) 59 75 Boiler Elec. (kWh) 14 16 Total Input (Btu/ kWh) 5,857/1.71 6,002/1.76 Output to Zones (Btu/ kWh) 3,801/1.11 3,702/1.08 Total System Gas Input (Btu/ kWh) 6,408/1.88 6,502/1.90 Total System Elec. (kWh) 44 50 Total Input (Btu/ kWh) 6,558/1.92 6,674/1.95 Total Output (Btu/ kWh) 5,493/1.61 6,018/1.76 Overall System Efficiency 66.8% 67.0% Space Heat Efficiency 72.5% 69.8% Boiler's Heating Efficiency 95% 99% Total Cost $0.095 $0.098 Control Settings Constant Setback Temperature Space Heat 6,274/1.84 5,251/1.54 80 52 18 11 6,441/1.89 5,365/1.57 3,999/1.17 3,492/1.02 Total System 6,935/2.03 5,929/1.74 53 40 7,116/2.08 6,066/1.78 6,309/1.85 5,116/1.50 66.6% 67.2% 69.7% 73.2% 98% 96% $0.104 $0.088
|Printer friendly Cite/link Email Feedback|
|Author:||Arena, Lois B.; Faakye, Omari|
|Publication:||ASHRAE Conference Papers|
|Date:||Dec 22, 2014|
|Previous Article:||Conversion of an Existing Transverse Ventilation System to a Longitudinal System Using a Saccardo Nozzle.|
|Next Article:||Typical Hot Water Draw Patterns Based on Field Data.|