Estimating Winter De-stratification Energy Savings in Large Warehouses
Richard Aynsley, Ph.D., M.ASHRAE Director Research & Development Big Ass Fan Company, Lexington KY, USA Abstract Various methods have been used to estimate the energy savings from de-stratification in large warehouses during the winter heating season. It is rare for these estimates to be compared with actual on-site measured data. A number of methods are examined and the associated assumptions are discussed and results evaluated. Only one of these methods is shown to give reasonable results. Introduction In recent years the cost of heating fuels used for heating large shipping and receiving warehouses have increased dramatically. This cost increase has stimulated interest in destratification as a means to improving heating energy efficiency. The simplest methods for estimating winter energy savings from de-stratification in shipping and receiving warehouses, focus on the difference in air temperature between floor level and underside of roof. They assume that after de-stratification the air temperature is uniform within the space. The estimate of energy savings is the difference between heat lost through the roof at the two temperature differences before and after destratification. This method ignores the influence of infiltration, ventilation or other sources of heat within the space, such as lighting, people and machinery. More complex methods attempt to account for these influences. The simple methods for estimating winter energy savings from de-stratification do not take account of heat loss by ventilation. Detailed tabular data on 30 year average outdoor winter temperatures and heating degree days for US and Canadian cities, referenced to 65° F, can be found in Chapter 24 of the 1981 ASHRAE Handbook of Fundamentals. Degree day methods are more fully described in Chapter 31 of the 2001 ASHRAE Handbook of Fundamentals. Heat losses due to ventilation and infiltration can amount to 20% to 50% of the total heat loss from a building. Not accounting for this heat loss due to ventilation in estimating energy savings from de-stratification can lead to significant errors. Heat losses due to ventilation can be reduced in some buildings by recovering heat from exhausted air and transferring the heat to incoming fresh make-up air using a heat exchanger. Shipping and receiving warehouses have numerous loading and unloading docks. These are difficult to seal and account for most losses of heated air, so heat recovery from ventilation is not generally feasible. What is needed is a method that is relatively simple but can produce estimates within say 10% of measured values using readily available data. Some estimate of the cost of fans and the energy used by fans to de-stratify the air is also needed to assess the cost benefit of installing and operating the fans.
1
Simplified Estimation of De-stratification Savings The following method for estimating savings due to de-stratification is widely used. For example take the 58,000 ft2 shipping and receiving warehouse in the State of New York. Gas consumed by the forced-air gas furnaces over an 89 day period from February through April, 2003 was 19,542 dekatherm. During these 89 days the heating degree days were 2220 over a base of 65 °F, and the average outdoor air temperature was 41.0 °F. The indoor thermostat temperature was maintained at 65 °F. Before destratification the air temperature at the underside of the roof deck 29.5 ft above floor level, was 89°F. The roof is estimated to have an average heat transfer coefficient of 0.13 Btu/hr.ft2.°F. The temperature difference before de-stratification, TDBD, is determined between the underside of the roof deck, measured on-site, and the average outdoor heating season temperature from climatic records for the location. For the example, this is (89 – 41.0) or 48.0 °F. The temperature difference after de-stratification, TDAD, is estimated between the underside of the roof deck, taken as the average of air temperatures at floor and roof deck, and the average outdoor heating season temperature from climatic records for the location. For the example, this difference is ((89 + 65)/2) – 41 or 36 °F. The heat lost through the roof before de-stratification, HLBD, is calculated using Equation 1. HLBD = C x A x TDBD (Btu/hr) = 0.13 x 58,000 x 48 = 361,920 Btu/hr Where HLBD = Heat lost through the roof before de-stratification (Btu/hr) C = Average Heat transfer coefficient for the roof (Btu/hr.ft2.°F) A = Area of roof (ft2) TDBD = Temperature difference through roof before de-stratification (°F) The heat lost through the roof after de-stratification, HLAD, is calculated with Equation 2 HLAD = C x A x TDAD (Btu/hr or W) = 0.13 x 58,000 x 36 = 271,440 Btu/hr Where HLAD = Heat lost through the roof after de-stratification (Btu/hr) C = Average heat transfer coefficient for the roof (Btu/hr.ft2.°F) A = Area of roof (ft2) TDAD = Temperature difference through roof after de-stratification (°F) The seasonal gas fuel saved, FS, due to de-stratification is calculated using Equation 3. FS = HSH x (HLBD – HLAD)/ (CV x E) = 2136 x (361,920 – 271,440)/(100,000 x 0.7) = 2761 dekatherm Equation 3 Equation 2 Equation 1
2
Where FS = Seasonal gas fuel saved due to de-stratification in dekatherm HSH = Heating season hours for the location CV = Calorific value for natural gas fuel, typically 100,000 Btu/dekatherm E = The efficiency of the gas furnace, typically 0.7 This gas fuel saved can be expressed as a percentage of the gas used before destratification or 100 x 2761/19542 = 14.1%. It should be noted that this method does not account for heat losses due to ventilation, or infiltration through the building envelope, or indoor heat gains from lighting, people and machinery. For large single storey buildings such as shipping and receiving warehouses, the roof surface typically offers the largest surface for heat loss. Also if indoor air is allowed to stratify, the indoor/outdoor temperature difference will be higher through the roof, than the average indoor/outdoor temperature difference across the walls. While the heat loss by conduction through the walls is relatively constant, the heat transfer by conduction through the roof can vary significantly. In winter, this is due mainly to the insulating effects of accumulated snow, wind, and radiant heat exchange with the sky. The later depends on cloud cover and the presence or absence of snow cover. Radiant heat losses can be very high on a winter night when winds are calm and the sky is cloudless. It is often difficult to assign a value for the thermal conductivity for envelope elements in existing buildings used in this method. Detailed construction drawings or specifications are rarely available, and even if they are they do not necessarily represent how the building was actually built. In any case roof insulation often varies in thickness in order to provide falls for rainwater drainage to outlets. Older insulation can be compromised by absorption of moisture, or by compaction of insulation over time. A Better Approach to Estimating De-Stratification Savings There is a better way to estimate winter heating savings due to de-stratification. Occupants of existing buildings usually have well documented data on the amount of gas used during a heating season from their utility company invoices. These invoices also often have the degree days of heating associated with each billing period. If heating degree data is not provided on invoices, it can be obtained from local climate records along with the average outdoor air temperature during a particular period of time. These data can be used to calculate the overall rate of heat loss through the building envelope over a given period. Equation 4 comes from Pignet and Saxena (2002). UA = H / (AIT - AOT) Where H= Rate of heat loss through an element of the building envelope (Btu/hr) U= Heat transfer coefficient for the element including air films (Btu/hr.ft2.° F) A= Surface area of the element (ft2) AIT = Average indoor air temperature (° F) AOT = Average outdoor air temperature (° F) The amount of heat released inside the building from other sources other than the furnaces, such as lighting, people and machinery, can be estimated as indicated earlier in Btu/hr.°F Equation 4
3
this paper. This is important because, during the warmer winter months, if the thermostat setting is maintained, heat from sources other than the furnaces will provide an increasing proportion of the heating to the building. Lighting In warehouses, lighting is likely to be the largest source of winter heat input apart from heating equipment such as furnaces. The amount of heat released by various types of lighting fixtures can be found in Chapter 29 of the 2001 ASHRAE Handbook of Fundamentals. Most warehouses use a combination of high-bay, high-pressure sodium discharge lighting fixtures, and lower level fluorescent fixtures on storage racks. Heat Gain from People Occupancy of shipping and receiving warehouses is often transient, with a small receiving crew operating fork lifts for much of the time, and large increases in occupancy at times when shipping crews come to load up orders of goods for their customers. While these people are working inside the warehouse they release heat into the space. By accounting for the numbers of people, their metabolic rate, and the average hours per day they are in the space, an average daily or weekly rate of heat release can be calculated. Values for heat released by people engaged in other activities are provided in Chapter 29 of the ASHRAE Handbook of Fundamentals under nonresidential cooling load calculating procedures for air conditioned space. Machinery Typical machinery used in shipping and receiving warehouses is limited to batterypowered forklifts. These machines have two electric motors, one for horizontal motion, the other for lifting the fork. The total electrical rating for both motors in a single machine is typically 7.4kW. Values for heat released by appliances are provided in Chapter 29 of the ASHRAE Handbook of Fundamentals Typical Aggregated Heat from Lighting, People and Machinery Aggregating the typical quantities of heat released from lighting 39.17 kW, people 1.30 kW, and machinery, 13.53 kW in shipping and receiving warehouse with a floor area of 58,000 ft2 gives a total of 54.00 kW. It is often convenient to express this time averaged heating energy in Btu/hr.ft2 of floor area. It should be noted that 1 W is equivalent to 3.41 Btu/hr. For the 58,000 ft2 this is equal to (54.00 x 1000 x 3.41499)/58,000 or 3.18 Btu/hr.ft2 for typical shipping and receiving warehouses. For example take the 58,000 ft2 shipping and receiving warehouse in the State of NewYork. Gas consumed by the forced-air gas furnaces mounted 22 ft above floor level, over a 89 day period from February through April, 2003 was 28,678 Therms. During these 89 days the heating degree days were 3432 over a base of 65 °F, and the average outdoor air temperature was 41.0 °F. The indoor thermostat temperature was maintained at 65 °F. Before de-stratification the air temperature at the underside of the roof deck 30.5 ft above floor level, was 89°F. In the method proposed by Pignet and Saxena, (2002), the estimated average indoor air temperature of stratified air assumed a linear vertical profile between floor and
4
underside of the roof. In this case, the average air temperature after thorough mixing would be (89 +65)/2 or 77 °F. The average seasonal rate of heat loss UA through the building envelope in Btu/°F before de-stratification can be calculated using Equation 4. The heat provided by the furnace during the sample heating season is 19,542 dekatherm times the calorific value of natural gas, 100,000 Btu/dekatherm, times the efficiency of the furnace. 0.7. This equals 1,367,940,000 Btu. Heat from other sources in the warehouse during the sample heating season is the heat release rate 3.2 Btu/hr.ft2, times the number of days in the sample heating season 89 days converted to hours, times the floor area of the warehouse, 58,000 ft2. This equals 396,441,600 Btu. From Equation 4: UA = (1,367,940,000 + 396,441,600) / (77 – 41.0) = 49,010,600 Btu/°F The reduced heat load, RHL, due to de-stratification is calculated using Equation 5. RHL = (UA) x (IATB – IATD) = 49,010,600 x (77 - 65) = 588,127,200 Btu Where RHL = Reduced heat load due to de-stratification (Btu). UA = Lumped time averaged rate of heat loss for the building envelope (Btu/°F). (for each degree of indoor/outdoor season average temperature difference) IATB = Average indoor air temperature before de-stratification (°F) IATD = Average indoor air temperature after de-stratification (°F) (= thermostat setting) This is converted to the equivalent quantity of gas EQG (dekatherm), by dividing by the calorific value of the gas per dekatherm and the heating system efficiency (0.7). EQG = RHL/CV x E Equation 6. = 588,127,200 / (100,000 x 0.7) = 8,402 dekatherm where EQG = Equivalent quantity of gas CV = Calorific value for the gas E = Efficiency of the heating appliance Expressing this as a percentage of the gas used before de-stratification, or 100 x 8,402/19,542 = 43%. Estimating Average Indoor Air Temperature A critical factor in de-stratification is that the indoor air is thoroughly mixed to an even temperature. When this is done, the difference between the temperature of air near floor level and at the underside of the roof deck is usually no more than 1 °F. Before destratification the difference between the temperature of air near floor level and at the underside of the roof deck can be up to 30 °F. 5 Equation 5.
Most simple methods for estimating winter energy savings from de-stratification use a linear profile for air temperature with height above floor level. In reality the air temperature profile is often not linear. The vertical profile of air temperature inside a space is strongly influenced by the distribution and height of heat sources in the building. Most shipping and receiving warehouses in the USA use a number of natural gas fired forced air furnaces mounted about 20 ft above floor level. On-site measurements were made of the air temperatures in a 33 ft high shipping and receiving warehouse with two forced air gas furnaces. These measurements indicated that air temperatures were constant to at least 5ft below the roof deck. This suggests that the vertical profile of air temperature was similar to a D profile. Average indoor air temperature in a space with a D type vertical profile of air temperature, after thorough mixing, can be estimated by taking the average of two temperature-weighted volumes of the indoor vertical profile of temperature above and below the horizontal part of the profile that aligns with the height of the furnaces above the floor. Whenever possible it is advisable to measure the vertical profile of air temperature on site.
Figure 1. Characteristic Vertical Profiles of Indoor Air Temperature Determined by Location of Warm Air Supply and Mixing (Andersen, 1998) A New Method for Estimating De-Stratification Savings In a departure from the method suggested by Pignet and Saxena, 2002, the writer suggests that averages of indoor air temperature, AIT, after thorough mixing should reflect the characteristics of the vertical profile of indoor air temperature. In the case
6
study, this can be done by using Equation 7, to account approximately for the D type vertical profile of air temperature. AIT = ((TAH x FA x HAH) + (TT x FA x HBH))/ (FA x (HAH + HOH)) Eq. 7 = ((89 x 58,000 x 8.5) + (65 x 58,000 x 22)) / (58,000 x (8.5 + 22)) = 71.69 °F Where AIT = Average indoor air temperature (after mixing) °F TAH = Air temperature above heaters °F FA = Floor area of facility ft2 HAH = Height above heaters to roof ft TT = Thermostat temperature setting °F HBH = Height below heaters to the floor ft The average seasonal rate of heat loss, UA, through the building envelope in Btu/°F before de-stratification can be calculated using Equation 5 where AOT is the average outdoor air temperature for the sample heating season. The heat provided by the furnace during the sample heating season is 19,542 dekatherm times the calorific value of natural gas, 100,000 Btu/dekatherm, times the efficiency of the furnace. 0.7. This equals 1,367,940,000 Btu. Heat in Btu from other sources in the warehouse during the sample heating season is the heat release rate 3.18 Btu/hr.ft2, times the number of days in the sample heating season 89 converted to hours, times the floor area of the warehouse, 58,000 ft2. This equals 393,963,840 Btu. Using Equation 4: UA = 1,367,940,000 + 393,963,840 / (71.69 – 41.0) = 57,409,705 Btu/°F The reduced heat load, RHL, due to de-stratification is calculated using Equation 5. RHL = 57,409,705 x (71.69 -65) = 384,070,926 Btu This is converted to the equivalent quantity of gas EQG (dekatherm), by dividing by the calorific value of the gas per dektherm and the heating system efficiency (0.7) using Equation 6. EQG = 384,070,926 / (100,000 x 0.7) = 5,487 dekatherm Expressing this gas fuel as a percentage of the gas used before de-stratification, 100 x 5,487/19542 = 28%. The dollar value of savings over this sample heating season are 5,487 x $0.8 = $4390. The sample heating season used of 2,220 HDD to a 65°F base cannot be representative of longer term heating seasons for the local climate. An estimate can be made of the savings over a typical heating season, based on 30 year climatic data by expressing the heating demand for a heating season in terms of dekatherm/HDD. In this case the 30 year ASHRAE design season for the location is 4871 HDD to a 65°F base. These data are indicated for cities in the USA in the 1981 ASHRAE Handbook of
7
Fundamentals. Estimated savings for the ASHRAE heating season based on 30 year data are $4,390 x 4871/2220 = $9632. Comparison of Estimated Energy Savings with Measured Savings The data used in the two estimates detailed above relate to a 58,000 ft2 shipping and receiving warehouse in Middletown NY. The owners of this facility collaborated in a study to evaluate the cost savings of de-stratification in winter. Gas consumption by the heaters along with heating degree-days (base 65°F) were recorded for February, March, and April in 2003 before de-stratification fans were installed. Similar measurements were made during the same months in 2004 after de-stratification fans were installed. These data are shown in Table 1. The other facility related data used in this paper were provided by the facility owners. Table 1. Gas Used and Heating Degree Days at the Facility Month
Gas Used In 2003 Before de-stratification (dekatherm) Gas Used In 2004 After de-stratification (dekatherm) HDD in 2003 Heating Season HDD in 2004 Heating Season
Feb. March April Totals
Average
10,358 5,798 3,386 19,542 6514
6,259 4,201 2,242 12,702 4234
1,011 736 473 2,220 740
889 698 373 1.960 653.3
Given that each winter season is unique, direct comparisons cannot be made. To overcome this, one can compare the quantity of fuel required to provide the same indoor condition divided by the severity of the winter season in heating degree days. In this case for 2003 before de-stratification, this equals 19,542/2,220 = 8.8 dekatherm/HDD. In 2004 after de-stratification, this equals 12,702/1,960 = 6.48 dekatherm/HDD. This represents a 26.4% reduction in gas used. The first method for estimating winter energy savings from de-stratification gave a saving of 14.1%, This was 46.6% less than actually measured in the case study. The second method by Pignet and Saxena, gave a saving of 43%. This was 62.9% more than actually measured in the case study. The third method by Pignet and Saxena, modified to better account for the vertical profile of indoor air temperature for gas furnaces mount on the walls gave a saving of 28%. This was 6.1% more than the measured value. This is within 10% of the measured value, a criteria often taken as acceptable for engineering estimates of this type given the limitations of the broad assumptions used.
Cost of Operating Large Ceiling Fans for De-Stratification Energy-efficient, large, low-speed ceiling fans are purchased principally for destratification in winter. A 58,000 ft2 shipping and receiving warehouse would typically
8
have five 20ft diameter fans operating quietly at approximately 20 rpm continuously. While providing the de-stratification the air velocity at head-height above the floor is kept below 50 fpm to avoid drafts. Cost of purchase and installation of the five large industrial ceiling fans would be approximately US$23,000. At 20 rpm, each fan would use 0.1 kWatts of power. Over the 89 day period of this study these fans would use 5 x 0.1 x 89 x 24 or 1068 kWhrs of electricity, costing approximately 6c per kWhr, or $64.08. These fans can be operated up to a speed of 49 rpm for cooling air movement in summer. Conclusions Three methods for estimating the winter energy savings from de-stratification were compared with field data for a three month sample heating season. The first method was extremely simple but underestimated the actual savings by 46.6%. The second, more detailed engineering approach was developed by Pignet and Saxena (2002). This method failed to account for the vertical profile of indoor air temperature developed when gas furnaces are mounted high on the walls. This method overestimated the energy savings by 62.9%. The third method, was the Pignet and Saxena method, modified by the writer to account for vertical profiles of indoor air temperature. This method overestimated savings by 6.1%, but met the criteria of being within 10% of actual measured value. The benefit of the Pignet and Saxena (2002) method is that it computes a lumped heat transfer coefficient for the building from seasonal heat loss based on heating energy used. This avoids difficulties in measuring, or otherwise determining, the heat transfer coefficient for the building envelope and other losses due to infiltration and ventilation. The modification suggested by the writer to the Pignet and Saxena method, better takes into account the influence of the type of vertical profile of indoor air temperature on the average indoor air temperature. Winter conditions vary from year to year. This complicates year to year comparisons of the efficiency of heating energy use. To overcome this difficulty it is suggested that comparisons between years be made on the basis of fuel used per heating degree day when comparing fuel used per heating season. REFERENCES Andersen, K. T. (1998) Design of natural ventilation by thermal buoyancy with temperature stratification. Proceedings of 6th International Conference on Air Distribution in Rooms, ROOMVENT ’98, Vol. 2, pp. 437-444. ASHRAE (1981) Handbook of Fundamentals, Chapter 24, American Society for Heating Refrigerating and Air-conditioning Engineers, Tulie Circle NE, Atlanta. GA, ASHRAE (2001) Handbook of Fundamentals, Chapter 29, American Society for Heating Refrigerating and Air-conditioning Engineers, Tulie Circle NE, Atlanta. GA, Pignet, Tom and Saxena, Umesh (2002) Estimation of energy savings due to destratification of air in plants, Energy Engineering, Vol. 99, No. 1, 69-72.
9