MILDURA RIVER HOUSE: PERFORMANCE A “LUXURY” SUSTAINABLE HOME

Luke MIDDLETON1 1 EME Group, Middle Park, Melbourne, Australia, luke@emegroup.com.au

Oral presentation preferred

Keywords: sustainable housing, thermal performance measurement, thermal performance computer modelling, thermal mass vs. insulation

Abstract

This is a twelve month monitoring project; this paper will present the first five months of measured data of environmental conditions in new luxury home and compare this with computer simulated predictions of two types of software: the housing industry standard rating tool, First Rate and the more sophisticated modelling software EnergyPlus.

The aim for this house is to reduce energy consumption by at least 50%, improve air quality and provide significant reductions in mains water usage. The premise is that intelligent efficiency is holistic focusing on an interaction between site, people, landscape, building envelope, seasons, activities and orientation. Ecological sustainability can only succeed when underpinned by an active participation with these systems, and this project sets out to encourage and make this participation convenient and enjoyable.

This project addresses the challenge of achieving sustainable function in a 'luxury' home. This is a timely opportunity as more people aspire to the costly environmental qualities of size and grandeur. This project is a symbiosis of sustainable design with a luxurious living, underpinned by embedded environmental design components including: in ground Labyrinth cooling with night heat purging, insulated and non insulated rammed earth walls and low e double glazing with strategic window positioning.

This paper illustrates two major issues when using energy modelling software. Firstly, regarding the 'FirstRate' software, the actual building performance is significantly better than what was predicted by the software, which rated it as 5.5 Star. Secondly, regarding the more sophisticated simulation software EnergyPlus, the actual performance still outstrips the predictions.

1. Introduce project

The project is located in the township of Mildura on the banks of the Murray River. The climate is semiarid, experiencing large fluctuations between day and night time temperatures. In summer the temperatures can reach in excess of 42C, whilst in winter sub-zero temperatures are experienced overnight. On average throughout the year the night time temperature can be 15-20C below the day time temperature. (figures obtained from Special Services Unit of the Bureau of Metrology)

The clients brief was heavily steeped in Asian aesthetics and Feng Shui principles. A report prepared by a local Feng Shui consultant stipulated in detail the location and orientation of functions throughout the home. The obvious appeal of Murray River views were also a priority of the client. This required expansive views to the east thereby necessitating large areas of glass along the east elevation.

The design response was to provide large framed views to the east, reducing the area of window to less than 60%. Generous glazing was provided to the northern end of the building. Western and southern facades are massive insulated rammed earth walls with minimal glazing. Internally the public and private spaces were divided with a long rammed earth wall running north south, providing good internal thermal mass (^).

Other integral design components include, roof profiled to improve heat stratification, Labyrinth cooling with night heat purging, landscape design to provide mediation zones and shading, low energy diffuse lighting throughout and Rain water harvesting

Figure 1 Mildura House

1.1 Aims of the paper

This paper will present the performance of the home over five months. These results will be compared with the results predicted by computer modelling software. The paper will compare two types of computer simulation evaluation software, First Rate and Energy Plus with the actual performance of the home.

The aim is to demonstrate the limitations of computer simulated thermal performance for a luxury home with a creative non-typical design located in a semi arid climate. In particular, the performance of the various thermal mass objects, the labyrinth cooling system and thermal stratification.

With the advent of increased regulation regarding the performance of buildings, namely 5 star energy ratings in the instance of housing, there has been and will continue to be a significant growth in the use of computer modelling software. For real reductions in housing energy consumption it is essential to continually test against the actual performance of homes. This paper adds to previous research in this area such as the monitoring of the Ecohome at Cairnlea estate Melbourne between June 2004 and May 2005 prior to its occupation. The results of this research will assist in the improved understanding the computer climatic modelling software, it’s benefits and limitations. Without this vital auditing process inefficient homes and cities will continue to be built.

Furthermore, this research focuses on creative non-standard design solutions that cannot be realistically and fairly assessed by industry standard rating tools.

2. Data collection

2.1 Physical Monitoring

Sensors were obtained from Cornerstone Technologies and programmed to take temperature and humidity readings at half hour intervals.

Figure 2 Sensor locations

Half hourly measurements of temperature and humidity were obtained at strategic locations. Internally two sensors monitored the air temperature and humidity. As can be observed in Fig. 2, Logger D was located within the living space on the eastern wall to represent the peak temperatures that would be likely to be experienced within the space due to the large windows in the vicinity. Logger 13 was located in the predicted coolest zone of the house in-between the guest bedrooms. This zone has minimal glazing and is surrounded by the thermal mass of rammed earth walls. Sensor 1 was located outside above the bay window (2.6m metres above floor height).

LOGGER GENERAL SPECIFICATION

Center 340 thermorecorder -

Measurement Range: -30°C ~ 70°C ; -22°F ~ 158°F

Response time: 20minutes to 95% under slow air.

Accuracy: ±0.7°C(± 1.3°F)

Resolution: 0.1°C / 0.1°F

2.2 FirstRate software

The house energy rating is the method used to measure the energy efficiency of a house.

The house energy rating measures the energy efficiency of a house by allocating a point score to each element of the house e.g. floors, walls, windows, etc. and provides an overall rating on a scale from 0 to 6 stars, with half star increments.

An energy efficient house rates 5 stars or higher. The house energy rating is independent of the size and type of housing. This means that large and small houses, attached and detached dwellings each have the potential to achieve a good energy efficiency rating.

2.3 EnergyPlus software

The building was modelled in ECOTECT and outputted to EnergyPlus.

The model was constructed from detailed plans and elevations of the finished home. All as built materials were included in the model, including actual configuration of floor, wall and ceiling insulation, construction materials, window frames and double glazing units. The surrounding landscaping including significant river red gum trees to the east were modelled to emulate the shading of the building that occurs.

The Labyrinth Night Purge Cooling system was added in Energy Plus.

3. Results

3.1 Physical Monitoring

The results contained in this report are for the period 16 October 2007 until 26 March 2008. During the five month testing period the house was vacant during the period 10th - 25th November and 13th - 30th January.

The occupants were pleased with the comport levels for the majority of the time. During prolonged heatwave periods the backup air-conditioning system was used. Over the 162 day period of monitoring the backup air-conditioning was required for less than 6 percent of the time the building was occupied.

This summer period was unusually hot and there was one period of six consecutive days and another period of eight consecutive days with day time temperature reaching 40+ degrees Celsius (refer figure 3 and 4).

3.2 First Rate

The building was rated by Gary Kruithof of Dynamic Design. It achieved a 5.5 star rating with a score of 14 points including a deduction of 4 points from the Area Adjustment. The rating assumed Climate Zone 27. The winter score was 32, whilst for summer it scored -14.

Figure 3 Actual results for October, November and December 2007

Figure 4 Actual results for January, February and March 2008

3.3 Energy Plus

The EnergyPlus results predict that the building significantly mediates the extreme outdoor temperatures providing a comfortable environment for the majority of the summer months (refer figure 5 and 6). It can be seen in figure 5 and 6 that consecutive hot days and warm nights force the internal temperature up as the thermal mass has no chance of purging heat energy.

Figure 5 EnergyPlus results for October, November and December 2007

Figure 6 EnergyPlus results for January, February and March 2008

4. Discussion

The results clearly illustrate that the building performs better than the EnergyPlus modelling. Further it is obvious that the FirstRate software vastly under estimated the thermal performance of this home. Especially if you consider it’s performance against the Ecohome at Cairnlea estate Melbourne.

4.1 Analysis of results

During the October period the actual outside ambient temperatures, seen in Fig. 3, were on average 5 degrees higher than the weather file inputs for the EnergyPlus model that can be observed in Fig. 5. Despite this, the measured internal temperatures remained within the comfort range. The EnergyPlus model predicted greater day to night temperature fluctuations and higher peak temperatures than actually occurred. During October there was no backup air-conditioning used.

A comparison of Fig. 3 and Fig. 5 also shows the simulated and actual results for November. For the majority of November the house was unoccupied and no night purging occurred. It can be observed that from the 13th-16th the increase in outdoor temperature triggered a steady rise in the internal temperature as there was no chance for thermal mass to cool through night purging.

The EnergyPlus model, in Fig. 5, provides for night time purging with the labyrinth from 8pm to 6am. When a period of similar outdoor temperature, is compared with the actual results it can be observed that despite the provision of night purging, the simulated indoor temperatures are higher than those observed in the actual results. For instance in the simulated results when the temperature rose to 36.5C outside the internal living room temperature was 31.6C whereas the actual results show that when the outside temp rose to 36.9 the internal living room temperature was 28.9C.

The results observed in December followed a similar pattern to October. Whilst the actual outdoor temperatures (Fig. 3) were far higher than the EnergyPlus weather file (Fig. 5), the internal temperatures were lower than were predicted by the computer model. It should be noted that during the excessive and prolonged heatwave experienced at the end of December the occupants utilised the backup air-conditioning. The air-conditioning thermostats are located within the guest bedroom precinct (i.e. the cooler part of the building) and were set at 25 degrees to provide a temperature of 26-7oC in the living spaces.

4.2 Actual versus predicted comfort

It can be seen from these results that the passive systems provide a greater level of comfort than predicted by the modelling software. Within the zones surrounded by thermal mass day to night temperature fluctuations were reduced to less than 3 degrees (outdoor temperature fluctuation 20oC). In the living spaces where the design criteria necessitated significant areas of glazing the internal temperature variants was generally limited to less than 5 degrees. By contrast, the EnergyPlus model predicted the living spaces to fluctuate by up to 7.5 degrees.

Actual results show that during the majority of the monitoring period the temperature remained within comfort levels. The most significant increases occurred during periods when the house was unoccupied when no night purging occurred or on consecutive days of extreme heat where the temperature was over 35oC for a prolonged period as in Fig. 4 from 2nd-5th and 15th-19th February. During some of these periods backup air conditioning was used to obtain comfortable indoor temperatures. In March Mildura experienced 16 days over 35oC, 8 of which exceeded 40oC, this is the longest period that backup air conditioning was required. It is important to note that whilst the temperature exceeded comfortable levels on some days it was still several degrees cooler than external temperatures. For instance Fig. 3 shows that on 15th November when the outdoor temperature reached its maximum of 37.7oC it was only 29.6oC in the living room and only 25.5oC in the bedroom. On 16th when the outdoor temperature reached its maximum of 40oC it was 30oC in the living room and 26.6oC in the bedroom.

The measured results include extended periods when the building was not occupied. It can be seen during these periods that the thermal mass gradually increased in temperature, as it had no chance to purge the heat energy at night. Even during these periods the actual results are far better than the predicted EnergyPlus results (which also included for the operation of the labyrinth for heat purging over night).

4.3 Advantages and limitations of EnergyPlus

The complex nature of algorithms within the simulating software provides far more data about what is contributing to the outputs, than could have been obtained by physical measurements.

For example the contributions to the temperature of a zone can be broken down into solar gain, ventilation gain, conduction/convection gains, ventilation and so on. Simulation provides useful design insight, as parameters can be varied one at a time and the corresponding change observed in the data output.

The hypothetical (simulated) results are reliant on the accuracy of the inputs. They are also limited by the algorithms, which are approximations of physical circumstances. In some cases the modelling software outputs are limited as they provide more general information. For example, the EnergyPlus outputs are limited to average temperatures within a zone as opposed to temperatures at a specific point within that zone.

The EnergyPlus software cannot attempt to predict the habits of occupation that could improve or inhibit the performance of the building.

4.4 Advantage and limitations of physical monitoring

The measured data represents the true, all-inclusive physics of the situation, including the physics of the sensors (accuracy, calibration, lag etc). It shows the true measurement at that time and represents everything that is going on, much of which is likely to be uncontrollable and often unknown.

The physical monitoring of air temperature and humidity within a room or zone can include logging devices at various locations with in a zone close to windows such as at the ceiling level and at occupation level to provide a better understanding of real comfort levels. (Unlike the Energy Plus results which have only average temperature outputs)

A sensor may provide distorted outputs due to unplanned circumstances such as sunlight directly hitting the sensor or a heat source such as a computer being placed near the sensor (although careful analysis of the figures combined with common sense should reveal such anomalies.

The sensors provide raw data according to temperature and not a breakdown of the energy sources such as solar radiation, radiant heat and shade air temperatures that make up the resultant temperature.

4.5 Comparative modelling

There is some correlation between the actual performance and energy plus output, so, to discover a further level of insight we modelled the same building in lightweight brick veneer construction, no labyrinth and single glazed comfort plus glass (rating the building as 5 star). The actual building has double glazing with one layer being low e. This exercise revealed a massive difference in the predicted energy consumption between the as built home (5.5 star) and the alternative modelled (5 star) (refer Table 1)

Table 1 Predicted Energy Consumption of Home as Built (5.5 star) versus

Standard Brick Veneer Construction (5 star)

summary

heating cooling total energy

rammed earth 3,645,201 6,262,071 12,576,956

lightweight w comfort plus clear glass 10,417,755 11,850,640 22,268,395

saving in energy 7,051,662 4,056,098 11,107,759 watt hours per year

percentage reduction in energy 293.5% 145.4% 188.3%

saving in energy with earth tube 4,398,410 watts per yr

5. Conclusions

It is clear from the measured results that the computer modelling of passive building design has significant limitations. The effectiveness of thermal mass and insulated thermal mass does not seem to be accessed effectively by First Rate and even the more detailed modelling capacity of EnergyPlus does not accurately predict the actual performance of this home.

Moving beyond the success of this project, the ongoing monitoring aims to inform and improve the occupants’ operation of the building. With more detailed understanding of specific temperatures throughout the house and within the labyrinth and rammed earth walls, the occupant can make informed decisions about the opening and closing of windows and the operation of the labyrinth duct system. From a design point of view these results are invaluable in consolidation of the original principles and will provide information to improve similar design approaches.

Further studies include: additional temp loggers to understand temperatures in and along labyrinth, temperature within the rammed earth wall versus surface temperature, and air temperature adjacent to better understand thermal lag and the effect of air temperatures versus solar radiant energy on the thermal mass, and internal temperature loggers at high and low level to better understand the effectiveness of heat stratification within the roof form.

References

2004, FirstRate Manual Version 4, Sustainable Energy Authority Victoria.

Building Commission. 2008a, 5 star standard of all new homes, http://www.buildingcommission.com.au/www/default.asp?casid=2797, accessed April 2008.

Building Commission. 2008b, 5 star- what to do now and for the future, http://www.buildingcommission.com.au/asset/1/upload/5_Star_Booklet_final_LR.pdf, accessed April 2008.

Bureau of Meteorology. 2008, Temperature and Humidity readings for Mildura, http://www.bom.gov.au/climate/, accessed April 2008.

Kruithof, Gary. 2006, FirstRate Report for 274 Cureton Avenue, Mildura.

National Renewable Energy Laboratory. 1995, User's Manual for TMY2s, http://rredc.nrel.gov/solar/pubs/tmy2/PDFs/tmy2man.pdf, accessed March 2008.

Rahman, S M S, Silvia, S D, Patnaikuni, I. 2006, ‘Housing Sustainability in Australia’, Proceedings of the International Symposium on Environmental Sustainability 2006, Dhaka, Bangladesh, 7-9 February 2006.

Rahman, S M S, Silvia, S D, Patnaikuni, I. 2005, “Instrumentation and monitoring strategies in

measuring the performance of sustainable housing”, in Collaboration and Harmonization in Creative

Systems, Hara (ed.), Taylor & Francis Group, London, pp. 269-274.

Simulation Research Group. 1996, ‘New Data Visualization Tools for DOE-2.1E Calibration’, Building Energy Simulation User News, Vol. 17, No. 2, pp 2-6, http://simulationresearch.lbl.gov/dirun/1702.pdf, accessed April 2008.

U.S. Department of Energy. 2007a, EnergyPlus Energy Simulation Software: Testing and Validation, http://www.eere.energy.gov/buildings/energyplus/testing.html, accessed April 2008.

U.S. Department of Energy. 2007b, EnergyPlus Testing with Building Thermal Envelope and Fabric Load Tests from ANSI/ASHRAE Standard 140-2004, http://www.eere.energy.gov/buildings/energyplus/pdfs/energyplus_ashrae_140_envelope.pdf, accessed April 2008.

xEsoView. 2006, xEsoView file viewer for EnergyPlus .eso output files, http://xesoview.sourceforge.net/, accessed April 2008.