811-584

8/11/2019 811-584

1/8

THE BENEFITS OF 8760HOUR-BY-HOUR BUILDING

ENERGY ANALYSIS

By: James PeguesSenior HVAC Systems Engineer, Carrier Software SystemsCarrier CorporationSyracuse, New York

April, 2002

8/11/2019 811-584

2/8

(Intentionally left blank for position only)

8/11/2019 811-584

3/8

IntroductionAs energy costs rise, building owners are

becoming increasingly interested in operatingcosts and energy efficiency. As a result, buildingenergy analysis (BEA) is becoming an importanttool in the HVAC design field. Currently manyBEA tools are available to engineers. Most are in

the form of computer programs and employ avariety of methods with different benefits.Among these, BEA tools such as Carrier's HAP

program that use the 8760 hour-by-hour methodcan offer the greatest benefits because they yieldhighly accurate, sophisticated systemcomparisons. This article will discuss the

benefits of 8760 hour building energy analysis by first explaining the basics of building energyanalysis and the requirements for high qualityBEA system comparisons. Then, major BEAmethods will be evaluated with special emphasison the benefits of 8760 hour-by-hour versus

reduced hour-by-hour methods.

What Is Building EnergyAnalysis?Building energy analysis is the technique ofestimating energy use and operating costs for a

building and its energy consuming systems. Inour industry, particular emphasis is placed on theenergy use of a building's HVAC systems. The

purpose of BEA is to compare the energy useand operating costs of alternate system designs

in order to choose the optimal design. Theanalysis mathematically simulates the thermal

performance of the building to determine coolingand heating loads. It then mathematicallysimulates the performance of HVAC equipmentin response to these loads to determine energyuse over the course of a year. Finally, energydata is used to calculate operating costs.

Requirements For HighQuality Results.Successful building energy analysis relies onconsidering as many of the physical factorsinfluencing building loads and equipment

performance as possible. The ultimate result ofthe analysis is a predicted operating cost. Anaccurate cost prediction relies on energy usedata, which in turn relies on equipmentsimulations, which must be based on buildingload predictions, all of which must be accurate.

Concisely stated, high quality results can beobtained when the analysis considers:

1. The Range and Timing of WeatherConditions. Varying levels of temperature,humidity and solar radiation during the yearinfluence building loads and equipment

performance. In each geographical location

conditions range from hot to cold, wet to dryand sunny to cloudy in different ways.Considering the actual ranges of theseconditions and when they occur on a daily,monthly and yearly basis is crucial to

producing accurate energy use results.

2. The Hourly and Daily Variation inInternal Loads. Patterns of building useinvolving occupancy, lighting andequipment operation can changesignificantly from one day to the next.Considering these use patterns in their

correct day-to-day sequence is important ingenerating accurate load data.

3. The Dynamic Nature of Building HeatTransfer. The process of converting heatgains and losses to cooling and heating loadsis a transient rather than a steady-state

process. Heat gains occurring during onehour often affect loads over a number ofsucceeding hours. Consequently, it isimportant to consider accurate sequences ofheat gains occurring during the day. Inaddition, because weather conditions and

building use profiles vary from day to day,sequences of heat gains can affect loadsfrom one day to the next.

4. The Response and Performance of HVACEquipment. How controls, systems andequipment respond to demands for coolingand heating in a building, and the factorsthat affect part-load performance of theequipment must be considered to yieldaccurate equipment energy use data.

5. The Details Of How Utilities Charge ForEnergy Use. Often prices for energy vary byseason and time of day. Further, charges areoften made for peak energy usage. As aresult, the analysis must not only be able to

produce accurate estimates of how muchenergy is used, but must also accuratelydetermine when during the day energy isused.

The Benefits of 8760 Hour-By-Hour Building Energy Analysis1

8/11/2019 811-584

4/8

Evaluation of BEA MethodsA wide variety of building energy analysismethods are currently available to HVACengineers and range from simple tosophisticated. The simplest methods involve thelargest number of simplifying assumptions andtherefore tend to be the least accurate. The most

sophisticated methods involve the fewestassumptions and thus can provide the mostaccurate results. Generally, BEA methods aredivided into three categories:

a) Single Measure Methods (example:Equivalent Full Load Hours)

b) Simplified Multiple Measure Methods(example: Bin Method)

c) Detailed Multiple Measure Method(example: Hour by Hour)

While methods in the first two categories serve a

useful role in providing quick, preliminaryenergy estimates, the simplifications theyinvolve impair their accuracy. Each will be

briefly discussed below. The main focus of thefollowing discussions, however, will be thedifferent hour by hour methods contained in thethird category.

Single Measure MethodsThese methods involve one calculation of annualor seasonal energy use. The Degree-DayMethod, for example, computes energy use bycombining one degree-day weather value with aload value and an efficiency value to obtainseasonal or annual energy use. Similarly, theEquivalent Full Load Hour Method combinesfull load capacity, full load efficiency andequivalent full load hours to obtain annualenergy use. In both cases, this level of simplicityis achieved by using such sweeping assumptionsthat the accuracy and reliability of these methodsare very limited.

Simplified Multiple MeasureMethodsThese methods involve calculations of energyuse at several different conditions. With the BinMethod, for example, energy use is computed ata series of outdoor air dry-bulb conditions.Results are then weighted according to thenumber of hours each dry-bulb condition is

expected to occur to determine annual energyuse.

For example, the temperature 47 F would beused to represent the range of conditions between45 F and 50 F, referred to as a "bin". Buildingloads and equipment energy use would first becalculated for the 47 F bin. Next, energy results

would be multiplied the number of hours peryear temperatures are expected to occur between45 F and 50 F to determine annual energy use forthat bin. Similar calculations would then berepeated for all other temperature bins for thelocal climate and would be summed to determineoverall annual energy use.

While the Bin Method provides a vastimprovement in sophistication over singlemeasure methods, it has a fatal flaw. This flaw isthat it must decouple weather conditions, loadsand system operation from time. For example,

hours in the 47 F bin, when the outdoor dry-bulbis between 45 F and 50 F, occur at a variety oftimes of day and night, days of the week andmonths of the year. Because a single calculationis performed to represent energy use for all thesedifferent times it is difficult or even impossibleto accurately:

a) Link solar radiation and humidity conditionsto the bin.

b) Consider hourly and daily variations ininternal loads.

c) Consider the transient hour to hour and dayto day thermal performance of the building.

d) Predict time-of-day energy use and peakdemands.

Inevitably averaging assumptions must be madeto shoehorn all these considerations into theframework of the bin analysis. And theseassumptions impair accuracy. While the BinMethod is useful for simple, preliminaryestimates of energy use and operating cost, itcannot provide the level of accuracy andsophistication offered by the detailed multiplemeasure methods.

Detailed Multiple MeasureMethodsThese methods perform energy calculations onan hour-by-hour basis. As a result, they have the

potential to satisfy all the requirements listedearlier for high quality energy analysis results.There is, however, a certain amount of variation


8/11/2019 811-584

5/8

among different detailed multiple measuremethods, leading some methods to meet theaccuracy requirements better than others. Withinthe detailed multiple measure category are twomajor sub-categories worth discussing:

a) The Reduced Hour-By-Hour Method b) The 8760 Hour-By-Hour Method.

Reduced Hour-By-HourMethod: How & Why?This method typically uses one 24-hour profileof average weather conditions per month. Energysimulations are performed for this average

profile and results are then multiplied by thenumber of days in the month to obtain monthlyenergy totals. Upon this basic foundation,different reduced hour-by-hour methods makevarious improvements to enhance the accuracyof results:

a) Some methods analyze building operationfor a typical Weekday, Saturday and Sundayeach month since building use profiles differsignificantly between these days. Oneaverage weather profile is still used for allthree typical day simulations each month.

b) Some methods also analyze equipmentoperation for a hot and cold day each monthin an attempt to improve estimates of peakelectrical demand.

c) Some methods simulate building operationfor one 7-day week each month to try toaccount for day-to-day building dynamics.However, a one average weather profile isstill used for all 7 days of the simulation.

The fundamental principle of this method is that building and equipment performance on hotterand colder than normal days each monthaverages out so that monthly energy use can beaccurately predicted by simulating a small groupof days using average weather conditions. The

method offers the benefits of reduced calculationtime and more moderate demands on computermemory and hard disk storage space.

8760 Hour-By-HourMethod: How and Why?This method simulates building and equipment

performance for all 8,760 hours in the year using

the proper sequence of days and actual weatherdata. No weighting of results or simplificationsare necessary.

The fundamental principle is that the way to produce the most accurate energy and operatingcost estimates is to mimic the real-time operatingexperience of a building over the course of a

year. All the requirements listed earlier for highquality energy analysis results can be met withthis approach. The actual weather data accountsfor the range and timing of weather conditions ingreat detail. Further, the hourly and dailyvariation of building occupancy, lighting andequipment use can be easily accounted for. Inaddition, the full year simulation tracks thedynamic hour-to-hour and day-to-day thermal

behavior of the building, and the response ofHVAC equipment to this behavior. The ultimateresult is high-quality data that can be utilized to

produce accurate, detailed data about the

quantity and timing of energy use. Both arerequirements for accurate operating costestimates.

Comparison of Reduced and8760 Hour-By-HourMethodsWhile the Reduced Hour-By-Hour method often

provides accurate results, the 8760 Hour-By-Hour method can consistently provide superioraccuracy and reliability. Among the reasonswhy, five stand out:

1. Better Estimates of Monthly Energy Use. One of the flaws in the fundamental principle ofthe reduced hour-by-hour method is that it relieson building and HVAC system behavior being"continuous" and "linear" during the month.Often these requirements are not met and thisadversely affects accuracy

"Continuous" is a mathematical term that in thiscontext refers to a consistent mode of operation(it does not refer to constant 24-hour operation of

equipment). For example, during a summermonth HVAC system operation is continuouswhen cooling is consistently done on all dayswhether hot, cool or average weather conditions

prevail. In this situation simulation of system performance for one 24-hour average weather profile has the best chance of approximating thetotal energy use for a month.


8/11/2019 811-584

6/8

However, system operation is often not"continuous" during a month, especially duringintermediate seasons. For example, during anautumn month, cooling may be done on warmerthan average days, economizer operation mayoccur on average days, and heating may occur oncooler than normal days. A simulation of oneaverage day for such a month may indicate little

or no cooling and heating because it does notconsider the warmer and colder than averageconditions.

In moderate climates this problem can becomesevere when the only time heating occurs is oncolder than average winter days. Because onlyaverage winter weather is considered, most or allof the heating duty for the year may be missed

by an average-day simulation approach.

Finally, "linear" behavior is a requirement foraveraging to be accurate. For example, if cooling

loads are 20% larger when the temperature is 15F warmer than average, and 20% smaller whenthe temperature is 15 F cooler than average,cooling loads are linearly proportional to outdoortemperature. Averaging of the warm day andcool day loads will result in loads similar tothose produced by simulating only the averageweather day. However, loads depend on morethan just outdoor temperature. Solar radiation,internal loads and hour-to-hour and day-to-daydynamic behavior also affect loads and oftenresult in non-linear behavior.

Another example involves cooling equipment. Ifequipment input kW decreases 8% for every10% drop in part-load ratio, input kW and loadare linearly proportional. If this relationshipholds true, simulation of equipment performancefor one average day per month has the bestchance of accurately approximating equipment

performance on the collection of hot, averageand cool days during the month. Unfortunately,the performance of equipment is often non-lineardue to part-load, entering condenser temperatureand other performance factors. Consequently, theaccuracy of the average day approach can bedegraded when equipment behavior is non-linear.

The 8760-hour method avoids these problems bysimulating building and equipment operation forthe entire month. Actual weather data used bythe simulation consists of a collection of days, allwith different combinations of temperature,humidity and sunshine.

Figures 1 and 2 provide an example of this kindof data for the month of September in Chicago.Figure 1 demonstrates the way dry-bulbtemperatures can vary during a month. Thedotted lines in this figure are the upper and lowerlimits of the average temperature profile for themonth that would be used by the reduced hour-

by-hour method. Comparison of the average and

actual data shows a significant number of hoursoutside the range of conditions considered by theaverage day simulation approach.

Figure 1 . Chicago Weather / September Dry Bulbs

Figure 2 . Chicago Weather / September Solar

Likewise, Figure 2 demonstrates the variation ofsolar radiation profiles during the month. Thedotted line indicates the maximum solar flux inthe average day profile used by the reducedhour-by-hour method. Once again, there aremany conditions with greater sunshine and lesssunshine than considered by the average dayapproach. More importantly, comparison of the

peaks and valleys in Figures 1 and 2 shows thathot days are not always sunny, and cool days arenot always cloudy. The diverse collection of hot,cold, sunny, cloudy and in between conditionsshown in these figures illustrates the complexnature of actual weather data and providesevidence that building loads will not be a simplelinear function of outdoor air temperature.


8/11/2019 811-584

7/8

By considering a diverse collection of weatherconditions each month, the 8760 hour method

produces a diverse, realistic set of cooling andheating loads for the month. Further, because afull month of days are simulated, the appropriatefactors influencing equipment performance areconsidered. There is no reliance on theassumption of "continuous" or "linear" behavior,

and the estimates of monthly energy use can behighly accurate.

2. Higher Quality System Comparisons.The issues discussed under item (1) affect notonly the accuracy of monthly energy estimates,

but also the quality of system comparisons. Thisis because many of the system designalternatives commonly considered exhibit

behavior that is both discontinuous and non-linear.

"Discontinuous" refers to inconsistent operation.

That is, operation that starts and stops rather thancontinuing for all operating conditions during amonth. "Non-linear" refers to the fact that thereis often not a simple proportional relationship

between load or outdoor temperature andequipment performance, as discussed in item (1).

A comparison of air handling systems with andwithout a non-integrated outdoor air economizer

provides a good illustration of this problem.With this type of economizer control,economizer dampers open when outdoor airtemperature drops below the supply airtemperature. The system can then immediatelyuse outdoor air for free cooling; mechanicalcooling can be turned off. For this example,assume the supply air temperature is 57 F andthat we are simulating system operation for theweather data shown in Figure 1. The dotted linesin this figure indicate the upper and lower limitsof an average temperature profile for the month.Because this average profile ranges between 58 Fand 75 F, a simulation using the reduced hour-

by-hour approach would never find a conditionwhen the economizer dampers opened duringSeptember; free cooling would never beavailable. However, with the 8760 hour method,the use of actual temperature profiles for themonth result in 119 hours during 13 days whentemperatures drop below 57 F. If cooling loadsexist during these times, the economizer wouldoperate to provide free cooling. Therefore,

because an economizer exhibits discontinuousoperation, turning on and off at specificconditions, the reduced hour-by-hour approachmay not be able to successfully account for its

operation. In our example the reduced hour-by-hour method would underestimate the benefit ofthe economizer.

Similar situations can exist for other systemcomponents and controls that involvediscontinuous, on/off behavior. Examplesinclude ventilation heat reclaim, supply air reset,

humidity control, cooling tower fan cycling, andloading and unloading of chiller networks aswell as many others.

3. More Accurate Load Histories.The fact that the 8760 hour method simulates

building thermal performance day to day for theentire year means it can correctly account for dayto day dynamic load behavior. This results inmore accurate load profiles, which ultimatelylead to more accurate energy use predictions.

For example, on a Monday morning in the

summer, pulldown loads tend to be larger thanon other days of the week due to the heataccumulated by the building mass during theweekend. In addition to resulting in largercooling loads, these conditions can sometimesset the monthly electric demand.

In reduced hour-by-hour methods that do notsimulate a full week of operation, each day issimulated separately from all other days.Consequently, day to day building dynamicscannot be considered, and the building loadhistories are more simplistic. In those reducedhour-by-hour methods that do simulate a 7-daysequence each month, results tend to beunrealistic since the same average weather

profile is used for all 7 days.

4. Higher Quality Time Of Use EnergyData.

Because of the diverse weather conditions andoperating conditions, and because of the dynamicnature of building heat transfer, 8760 hourmethods can produce energy use data that notonly accurately defines how much energy isused, but when during the day and week theenergy is used. When energy prices vary withtime of day, accuracy of the timing of energy useis critical for producing accurate operating costdata.

5. More Accurate Estimates of PeakDemand.

Finally, by considering the full range of weatherand operating conditions experienced by a

building during a month, the 8760 hour method


8/11/2019 811-584

8/8

is able to produce more accurate estimates of peak energy demand. When utility rates includea demand charge component, a significant part ofthe energy cost can be due to the peak energy userather than the quantity of energy used. Manyreduced hour-by-hour methods must determinedemands from average day simulations whichtend to underestimate demand values. Those

methods that add consideration of hot and coldday weather profiles can provide animprovement in demand estimates. However, itis important to recognize that demand will bedependant on more than just the outdoor airtemperature. Solar radiation, internal loads,

building use profiles and day to day buildingdynamics also play an important role. The 8760-hour method is the only method that cansimultaneously consider all these factors.

Thus, while the Reduced Hour-By-Hour Methodconsiders many of the factors required for high-

quality energy estimates, certain aspects of themethod are flawed and can limit the accuracy ofthe method. Because the 8760 hour method usesa more detailed, comprehensive approach to

building simulation, it can consistently overcomethese problems to provide accurate, reliableenergy estimates.

ConclusionThis article has discussed the important benefitsof the 8760 hour-by-hour building energyanalysis method. This method is certainly notnew. Computer programs using this method have

been available for nearly three decades.However, because many of these programs weredeveloped on mainframe computers as researchtools, the programs, and by association, the 8760hour method itself acquired a reputation of beingcomplicated, difficult and impractical to use. It isimportant to note that these are problems withthe implementation of the 8760 hour method, notwith the method itself. If the 8760 hour methodis implemented in a well-designed, well-documented microcomputer program, the 8760hour method can be as easy to learn and use asreduced hour-by-hour methods. In developingCarrier's Hourly Analysis Program we have puttwo decades worth of experience in the HVACsoftware field to work to produce an 8760 hourenergy analysis tool that is both powerful andeasy to use. The resulting program maximizesthe benefits of the 8760 hour method whileminimizing or even eliminating coststraditionally associated with use of this method.

www.carrier.com

Carrier Corporation Printed in USA 811-584

Documents

811-584