VA-89-2-4

Comparison of Freezing Control Strategies for Residential Air-to-Air Heat Recovery Ventilators E.G. Phillips, P.E. R.E. Chant, P.E.

' . . . ~

ABSTRACT A comparison of the energy performance of defrost

and frost control strategies for residential air-to-air heat recovery ventilators (HRV) has been carried out by using computer simulations for various climatic conditions. This paper discusses the results and conclusions from the comparisons and their implications for the heat recovery ventilator manufacturers and system designers.

INTRODUCTION Air-to-air heat exchangers with built-in fans and con­

trols (now called heat recovery ventilators or HRV) form an energy-efficient ventilation system in many new houses. In certain climatic regions, cold temperatures cause frosting of exhaust-side heat exchange surfaces. This reduces energy recovery in two ways:

1. Frost "fouls" heat exchange surfaces, reducing thermal conductance and heat transfer rates, and

2. Frost blocks exhaust airflow passages, reducing the energy available for recovery.

As a result, HRV manufacturers have developed strategies to prevent or limit the build-up of ice and frost in heat exchangers. However, little has been done to analytically compare the seasonal performance of these strategies under different climatic conditions.

ASHRAE T.C.-5.5-sponsored research project TAP 543, "An Investigation of Freezing Control Strategies for Residential Air to Air Heat Exchangers," had the following objectives:

• To analytically characterize freeze-control strategies used for residential air-to-air heat exchangers (HRV).

• To compare the need for, and effect of, various freeze­control strategies on the seasonal thermal performance of residential heat recovery units operating in different climates.

• To optimize specific freeze-control strategies and evaluate the associated energy savings.

This paper presents the results of this research project.

D.R. Fisher, P.E. B.C. Bradley, P.E.

APPROACH Work tasks in this project included an extensive

literature review, the development of a computer model to simulate HRV performance during freezing conditions, and the development of algorithms to simulate the energy per­formance of various frost control and defrost strategies.

A computer program based on findings from a literature study and first principles was developed for this study to provide a standard method to compare the energy performance of HRV freeze-control strategies under dif· ferent climatic conditions. The computer program con­sisted of three major components:

1. The HRV frosting model (a finite difference model of a counterflow heat exchanger to generate heat exchanger performance data vs. time at outdoor temperatures be­tween 50°F and -65°F (10°C and -54°C]).

2. The Erbs, Klein, Beckman (EKB) ambient temper­ature bin data generator (Erbs et al. 1983), which was us­ed to estimate the number of hours in temperature bins for selected locations.

3. Defrost and frost control algorithms. These used the HRV frosting and EKB model outputs to generate perfor­mance data for combinations of HRV, climate, and freeze­control strategy.

The HRV frosting model was run at 2°C temperature bins from 10°C to -54°C for HRV effectivenesses of 50%, 65%, and 80%. Data related to energy recovery rates and frost or ice accumulation (e.g., location, temperature, mass, thickness) were recorded at one-minute simulated time intervals for each temperature bin . The defrost and frost control strategy algorithms were applied to the H RV frost­ing model output and temperature data to predict the seasonal performance of the various effectiveness/climate/ strategy combinations. Variations of each freeze control strategy were simulated to assess the benefits of optimiz­ing the control strategy. Freeze control strategies are usually designed to ensure ventilation is maintained under extreme (i .e., cold) operating conditions. Optimization of control strategies was simulated by adjusting the f re-

E.G. Phillips and B.C. Bradley are Engineering Consultants with UNIES Ltd., Winnipeg, MB, Canada; R.E. Chant is Professor of Mechanical Engineering, University of Manitoba, Winnipeg; D.R. Fisher is a Consulting Engineer in Danville, CA.

THIS PREPRINT IS FOR DISCUSSION PURPOSES ONLY, FOR INCLUSION IN lsHRAE TRANSACTIONS 1989. V. 95, Pt 2. Not to be reprinted in whole or in part without written permission of the American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc .. 1791 Tullie Circle, NE, Atlanta, GA 30329. Opinions, findings, conclusions, or recommendations expressed in this paper are those of the author(s) and do not necessarily reflect the views of ASH RAE.

.. _ SEASONAL EF.FICIENCY

803 EFFECTIVE HEAT EXCHANGER c 100 T-- - - --------------. --e- RECIRC.

-- s.r.off __,,__ VAR.PH ..

--+-- RED c.: -- STG.BL

--+-- BLJ(.PH.

...

O +-----~----~-------l 5000 1 0000 . 1 5000 20000 ~

Fi~W~ 1

DEGREE D~YS (F')

Predicted seas.anal efficiency for an 80% effective HRIJ with various frost control stfategies operating in different climates

SEA.$0NAL EFFIC IEIK Y 65~ EFFECTIVE HE-AT EXCHANGER

100 ~--------------- -E-- R£CIRC

· - - sro~r

~ vMPH

§: BO

>­u z . ~ ' so S! ..._ u..

"' _J •O <( z 0

"' <(

"' 20 V1

0 5000

Figure 2

100

.g 80

>-u z ... 60 u ;;:: • ... "" _J 40 <( z 0 V1 <(

"' 20 V1

------ RED E.

~ SiGBL.

-+- BL~ PH

, 0000 1 5_09p 20000

DEGREE DAYS (Fl°

Predicted seasonal efficiency for a 65% effective HRV with various frost control strategies operating in different climates

SEASONA~ _EF_FICIENCY SD% EFFECTIVE HEAT EXCHANGER

--e- RECIRC.

-- s.r.orr _,,_ VAR.PH.

~RED, 1E~

-+- STG.BL.

--+-- BLK.PH.

.. ~ 0+---'----~----~------1 5000 10000 15000 20000

DEGREE DAYS (F) ' • ••• 1· •• ~

Figure 3 Predicted seasonal efficiency for a 50% effective HRV with various frost control strategias operating in different climates

quency, duration, 'oramount of frost control or defrost'used Vihen t.he system operated under less extreme conditions.

All freeze control strategies were evaluated using the same heat exchange(perforriiance and temperatUre pjp data. The use of common HRV and temperature data'for each strategy allows solid comparisons of the vcirious:· st_rategies even if the simulated heat exchanger perf6r-111a.nce data or bin temperatur~ data do not precisely . represent the performance of a particular HRV or the . ter!TPeratu re distribution of a given cl imate. ·

· (For a detailed description of the computer models, see Phillips et al. '1989.)

FROST CONTROL AND DEFROST STRATEGIES EVALUATED

In this pape( frost control refers to methods that pre­vent frost from forming in the heat exchanger, while defrost refers to methods that routinely remove frost from heat exchange surfaces. Jointly these are referred to as free;rn control strategies. In the study, freeze control strategies examined included ideal strategies and variations of strategies used in commercial/industrial and residential applications.

The following describes the frost control and defrost strategies evaluated in this study.

Ideal Freeze Control Strategy

The ideal control strategy continuotJsly removes frost from heat transfer surfaces without putting energy into the removal process. Thus heat exchanger fouling would not occur, anc;J the latent and sensible heat from cooling and free?ing WC1-ter vapor would be recovered. This represents the maximum amount of energy that could be recovered . in a heat exchanger of the specified effectiveness.

Ideal Less Ice Energy

This control strategy is similar to the ideal freeze con­trol strategy, except the energy required to melt frost and ice that has accumulateq. ir.t the heat exchanger core has been deducted from the energy recovered.

Defrost by Supply Fan Shutoff

Outdoor air supply is stopped while exhaust air con­tinues to flow uninterrupted. There is no heat recovery during defrost. Defrost is enabled when outdoor temper­atures fall below a predetermined temperature. The HRV operates in the defrost mode for set time periods (e.g., 5 minutes) at fixed ir:itervals (e.g., every 40 minutes) when defrost is enabled . This strategy is used in some produc­tion model HRV .

Defrost by Warm Air Recirculation

Ventilation is temporarily stoppep and house air is recirculated through the HRV to melt ice ·and frost deposits. The unit is operated in the defrost mode. for a fixed time period (e.g., 2 minutes) at fixed time intervals (e.g., every 20 minutes) whenever outdoor temperatures are, below, tre defrost setpoint. Airflow rates through the heat exchange,r must be increased to ensure the design ventilating ·rate is maintained, on average, during periods when the defrost

mechanism can operate. This increased airflow decrea~es heat exchanger effectiveness. . . .

This strategy is considered to be the state-of-the-art freeze-control method. Ii is used in seyeral residential HRV models.' '

Frost Control by Supply Air Preheat

Frosting is avoided by heating supply air above the frost threshold temperature before it enters the h~at exchanger. Block-switched and variable-output heater. controls were modeled. Block-switched controls change the preheater output in blocks (e.g.,· 1 kW), thus increasing supply air temperatures a fixed amount. Variable-output heater controls warm supply air to a temperature just above the frost threshold. )lo .

Supply air preheat is used in some residential air-to­air heat exchangers. Supply air preheat is also used in commercial and industrial applications.

Frost Control by Reducing Heat Exchanger Effectiveness

Heat exchanger effectiveness may be reduced by tilting heat pipes or by reducing the speed of. a rotary heat wheel. Whenever core temperatures fall below freezing, heat exchanger effectiveness is reduced so that exiting exhaust air is kept above its dew point.

This freeze control method is used in some commer-cial air-to-air heat exchanger applications. · ·

Frost Control by Reducing Supply Airflow Rates

The supply airflow through the heat exchp.nger is reduced during cold weather so exhaust air exit tempera­tures are kept above the dew point. The algorithms used for this strategy were the same as those used for reducing heat exchanger effectiveness.

This freeze control strategy is used in some commer­cial air-to-air heat exchange, appli~ations.

RESULTS AND CONCLUSIONS '1'

Figures 1, 2, and 3 plot' seasonal efficiency against heating degree data for var-ious fr~e.ze control strategies , and nominal HRV effectivenesses. The strategies plotted are:

warm air recirculation (RECIRC.) supply'fan shutoff (S.F.OFF) variable preheat (VAR.PH .) .... reduCing heat exchanger efficiency or reducing supply airflows (RED.E.) ·. staged block preheat (STG.BL.) fixed block preheat (BLK.PH.)

Figures 4, 5, and 6 plot the energy recovered against heating degree dayis data for the same freeze-control strategies an'd nominal heat exchanger effectivenesses. These figures also include a line for the ideal freeze con­trol strategy. ~n Figures 1, 2, and 3 the ideal would have a sea's'on'al effectiveness equal to the nominal HFJV effec­tiveness for all climates.)

ANNUAL ENERGY RECOVERED 80% EFFECTIVE HEAT EXCHANGER

12000---_:_- -----------, --e- IDEAL

~ 10000 :i: ;t ' ~ ,._ BODO (.? 0:: LU

~ 6000

c LU 0:: ~ 4000

0 u LU 0:: 2000

---- RECIRC.

-+- S.F.OFF

--- VAR.PH.

-+-RED. [.

-+- STG.BL

-.o- BLK.PH.

o.L------------------l 5000

Figure 4

1 0000 1 5000 20000

DEGREE DAY,S (F)

Predicted annual energy recovered for an 80% effective HRV with various frost control strategies operating in' aif-ferent climates 1

AMNUAL ENERGY RECOVERED . 65% EFFECTIVE HEAT EXCHANGER

12000----------------., --e- IDEAL

~ 10000 :i: ;t

" : ~000 (.? 0:: LU

~ 6000

c LU 0:: ~ 4000

0 u LU 0:: 2000

-- RECIRC.

-+- S.F.OFF

--- VAR PH

-+- RED. E.

-+- STG,BL.

-.o- BLK.PH.

o.1-------------------l 5000 10000 15000 20000

DEGREE DAYS (F)

Figure 5 Predicted annual energy recovered for a 65% effective HRV with various frost control strategies operating in different climates

ANNUAL ENERGY RECOVERED 50% EFFECTIVE -11EAJ EXCHANGER

12000---------...:...._ _____ --e- IDEAL

~ 10000 :i: ~

" >- BODO (.? 0:: LU z LU

c LU 0:: LU > 0 u LU 0::

6000

4000

2000

I --- RECIRC.

-+- S.F.OFF

-- VAR PH .

-+- REO. E.

-+- STG.BL.

-.o- BLK.PH.

o .1-..-------------~ 5000

Figure 6

10000 15000 20000

DEGRE£ OAYS (F)

Predicted annual energy recovered for a 50% effective HRV with various frost control strategies operating in different :.. climates .. '

"WARM AIR RECIRCULATION" STRATEGY EFF"ECT OF TIME INTERVAL BETWEEN DEFROST

100 ~---------------~ --e-- 21 MIN

g 80

>­u z •-••••--.,,._....,,.__..,._~.~~--~--_,, w eo • H "' u L;: u.. w

<i. 40 z 0 (/'J <(

~ 20

-- 15 MIN

EFF-B0%

EFF-~0%

0+-----~-----.--------4

5000 10000 15000 20000

DEGREE DAYS (F) ...

Figure 7 Predicted effect of reducing the time between defrost from 21 minutes to 15 minutes for the "warm air recirculation" defrost strategy

R'ECIRCULTATION VS IDEAL 1:.;· 803 -EFFECTIVE HEAT EXCHANGER

12000 ~--------------~ --e-- IDEAL

~ 10000 :i: ~

"" >- 8000 <:> 0:: w z w Cl w 0:: w > 0 (_) w 0::

6000

4000

2000

-- RECIRC

.,, 0+-----~----~-------1 5000 10000 - . 15000

DEGREE DAYS (F) 20000

., Figure 8 Comparison of the predicted annual energy recovered for

"the warm air recirculation" defrost strategy and the "ideal" strategy for an 80% effective HRV

"SUPPLY FAN OFF" STRATEGY EFFECT OF STRATEGY IMPROVEMENTS

· 1od~--------------~ --e-- STAGED

-- FIXED

g 80 r'4' A&

~~~====::::::2::~======::!i~ ~ - ..;c EFF-BO ...

I >-u ,,__ z w 60 Q LL u..

EFF-655'5

I w EFF-50% ...J 40 <( z 0 (/'J

< w 20 (/'J

0 5000 10000 15000 20000

DEGREE DAYS (F)

Figure 9 Co~~ari~oh iif seasonal efficien~y for staged (i.e., im­proved) and fixed defrost intervals for the "supply fan off" strategy

. ~ ,,.,,_.SUPPLY FAN OFF" vs IDEAL 803 EFFECTIVE HEAT EXCHANGER

12000 ~. IOEAL

--- STAGED

~ 10000 . --- FIXED

• :i: ····· ~ , ; ·' ~ >- 8000

<:> 0:: ' . w z 6000 w Cl w 0:: w 4000 > 0 u w 0::

20Q,~

0 5000 10000 15000 20000

DEGREE DAYS (F)

Figure 10 Comparison of the predicted annual energy recovereqfot . .an 80% effective HRV with "supply fan off" defrost strategy · "" to the ideal

Conclusions drawn from Figures 1 through 6 are:

1. Current state-of-the-art freeze control strategies function close to the ideal.

2. In climates with less than 7500 DD F (4200 DD C), the method of freeze control has iittle impact on seasonal performance, regardless of HRV effectiveness. In climates colder than 7500 DD F there are marked differences be­tween the seasonal energy performance of the various defrost strategies. ·

3. Greater energy savings are available in colder climates1 since lower temperatures occur, and for more hours. Thus the benefits of optimizing HRV freeze control strategies are greater in colder climates.

4. The higher the effectiveness of the HRV, the greater the energy benefits of improved frost control and defrost strategies. This is primarily because higher effectiv€1ness means greater amou.nts of energy a,r.e available for recovery at all temperatures. A second reason is that freez­ing conditions begin at higher supply air inlet temperatures as effectiveness increases; thus, in a given climate, the heat exchanger with the higher effectiveness requires its frost control strategy to be invoked for more hours.

5. Defrost by warm air recirculation, which appears to be the most efficient strategy in severely cold climates, is not the most efficient1in milder climates. The reason is the decline in HRV effectiveness due to the increased airflow rate is suffered at all times, while tt:ie benefits of this·strategy are realized only at low outdoorfemperatures.

6. In cold climates, preheating supply air to avoid frosting conditions significantly reduces recovered energy.

7. Reducing the heat exchanger· effectiveness or supply airflow to prevent frosting is an inefficient defrost strategy in cold climates. In temperature bins below the frost control setpoint, reducing supply airflow yielded the same energy recovery rate for all heat exchangers, regard­less of effectiveness. This is because the maximum energy that can be recovered is the amount that cools the exhaust-·. · air to the dew point.

Figures 7 and 8 'Illustrate the effect of reducing the time frame between defrost cycles fo"r the warm air recirculation

strategy. Reduced time frames between defrost cycles result in higher mean efficiencies, with the benefits being greatest in colder climates. However, the degree of improvement realized is small. Cutting the time between defrosrcycles from 21 minutes to 15 minutes increases the energy recovered by less than 1% in all cases examined. As initiating defrost results in energy losses not factored into the simulations and the cycling of components with finite operating lives, there are practical reasons to extend the time between defrost cycles. Figures 9 and 10 show that modest improvements in mean annual HRV performance can be realized with the supply fan shutoff defrost strategy by varying the duration of the defrost cycle with outdoor temperature. For the Arctic case, savings of up to 3% per year were p.r!=Jdicted by having defro!?t cycle durations of one, two, or three minutes (depending on outdoor temperatures) instead of a three-minute defrost cycle regardless of outdoor temperature. In milder climates the predicted benefits were smaller. · ·

Figure 11 shows preheater energy consumg.tion vs. recovered energy for each supply air preheat strategy for the 80% effective H RV. Staged block or variable preheaters are essential if preheater capacity exceeds 1 kW. Variable preheaters are significantly more energy effi­cient than fixed-output preheaters in colder climates.

Figures 12 and 13 show the impact of being able to reduce the setpoint'Of a variable preheater for the 80% ef­fective HRV. Enthalpy-type heat exchangers normally reduce the exhaust air humidity in cold weather, thus lower­ing the temperature at which frost control or defrost is needed. This could eliminate the need for defrost or frost control mechanisms in some mild climates. Reduced humidity would resu.lt in reduced frosting and exchanger surface fouling, thus higher energy recovery rates than for sensible heat exchangers. This fact (coupled with the find­ings that enthalpy heat exchangers reduce combined heating, cooling, and humidifying energy loads in most climates [Barringer and McGugan 1988]) strengthens the arguments for using enthalpy-type HRV.

.:.·

RECOMMENDATIONS I , • : .. _

1. Actual seasonal performance data from fi?ld moni- . toring are limited put are continui ng and shou ld be. -~ encouraged. However, to establish if ~the data are com pat- · ible, a survey of annual monitoring in pr:ogress should be made to standardize field testing methods with the aim of adopting consistent methods. . ,

2. A review of definitions used to..estab)ish the heat recovery. performance of HRVs is needf3d .... The many dif­ferent components that are being er'icornpas~ed by HRV systems are introducing new terminology to describe dif­ferent efficiencies which can be reduced to simpler terms.

3. Since .sever.al manufacturers are using crossflow cores, their performance and frosting conditions need comparing.ta the generic counterflow model studied in this project. ·· ···

REFERENCES

Barringer, C.G., and McGugan, C.A. 1988. "Investigation of en­thalpy air-to-air residential hi:iat exchangers.'' Final r,eport for ASH RAE Research Project 544-RP, September. , ·

~ ~ ~ >­(!)

15000

~ 10000 z LIJ

Q LIJ et:: LIJ > 0 u

5000 "" et::

o'd >-:::> Cl. ~

"PREHEAT" STRATEGY HEATER INPUT VS RECOVERED ENERGY FOR 807. EFF ./]-a- •All INPUT

-- VAR RECOV

_....,_ STG INPUT

-+- STG RECOV

• -+- FIX INPUT I -+-- FIX RECOV

10000 1 5000 20000

DEGREE DAYS (F)

Figure 11 Comparisons of the predicted annual energy recovered and preheater input energy for various preheat strategies for an 80% effective HRV

"VARIABLE PREHEAT" STRATEGY RECOVERED ENERGY AT VARIED sd POINTS

0

.' 807. EFF 1 0000 ' -a- 2s r

-- 19 r

~ 9000 _....,_ 1~ r

::c -+- s r -+-Jr

~ ~ >-

9000 (!)

• et:: LIJ z 7000 LIJ

Q w et:: w 6000 > 0 u LIJ et:: 5000

~oo+-~~~~~~~~~~~~~~--<

5000 10000 15000 20000

DEGREE DAYS (F)

Figure .J2 Comparison of the predicted annual energy recovered us­, ing various preheater setpoint temperatures for an 80% effective HRV

6000

z ~ 5000 ~ >-~ 4000 LIJ z UJ

5 3000 Cl. ~

~ 2000

3 ::c LIJ 1000 a:: Cl.

"VARIABLE PREH~AT" STRATEGY PREHEAT ENERGX AT VARIED SET POINTS, 803 EFF

" --8-;, 25 F

~19F'

0 4-~~--"=="'-.-~~~~~~~~~--l

5000 10000 15000 - 10000

DEGREE DAYS (F)

Figure 13 Comparison of the predi~ted an_nuafpreh_e_ater inpyt energy using various preheater setpq~r'it temperatures for an 80% effective HRV · '

Erbs, D.G.; Klein, S.A.; and Beckman, W.A. 1983. "Estimation of degree day on ambient temperature bin data from monthly average temperatures." ASHRAE Journal, June, pp. 60-65.

. ... .. ~

Phlllips. e.G.; Bradley, B.C.; Chant, R.E.; and Fisher, D.R. 1989. ''A model to compare freezing control strategies for re$iden­ti~I air-to-air heat recovery ventilators." ASH RAE Transa~t/ons, Vol. 95, Part 2 (In press) .