Home
The energy performance of a new or occupied building can be understood and explained in terms of the performance of its various systems, such as lighting and ventilation. An approach is described to achieve this understanding within reasonable time and cost. Two examples of detailed building assessments provide conclusions for design and management, and describe how these conclusions were arrived at. A headquarters office example shows how the approach can avoid a misleading assessment by separating non-building services loads and identifying critical building. A set of three retail premises shows how the approach can be useful when detailed yardsticks are not available for system energy consumptions, and indicates poor performance and significant potential for saving from a sales area lighting system.
The energy performance of an occupied building can be understood and explained in terms of the performance of its various systems, such as lighting and ventilation. An approach has been developed to achieve this understanding within reasonable time and cost.
The approach helps improve the quality of design and management of buildings by the use of performance benchmarks, and can allow high quality internal environments to be achieved with less operating cost and less environmental impact.
This paper describes briefly how this can be achieved and presents results from two examples of detailed building assessments with conclusions for design and management of the buildings, and describing how these conclusions were arrived at.
The examples show how the approach can be used to
* understand system performance,
* develop targets for the building and its systems,
* identify likely wastage,
* suggest remedial energy saving measures,
* support the development of customised energy consumption indices.
At the heart the method is the analysis represented by the tree diagram shown in Figure 1. In principle, data for all of the boxes in this diagram can be obtained for an existing building, or projected for a new building. The mē denominator measures building size, normally as treated or gross floor area depending on the benchmarks used, or as a sector-specific measure such as sales floor area for retail buildings.
At the top level, box A is calculated from the building's annual energy consumption which can be obtained from (or reconciled with) actual energy bills and provides an overall energy consumption index for comparison with published yardsticks [1,2]. Box A is the sum of all the boxes of type B, although if more than one fuel is used, each fuel should be assessed separately or combined in terms of a meaningful measure such as total cost or CO2 emissions, see reference 1.
Box B is an energy consumption index for each system, which can be compared with benchmarks published (for example) for offices [3], factories [4] and hotels [5]. Box B is boxes C and D multiplied together and divided by 1,000 to scale for kilowatt hours.
Box C is an index of installed power per mē which provides a simple but powerful check on energy efficiency of installed plant and equipment. It can be used directly [6] or combined to from a whole building index [7]. Box C is boxes E and F multiplied together (and in the case of lighting, divided by the scaling factor 100 because of the units used for lighting efficiency).
Box D is the effective hours at nominal full load for each system, which can be compared with typical values [3,8]. Box D is boxes G and H multiplied together.
Boxes E to H provide more detail for boxes C and D:
Box E is the level of service provided - the amount of light or ventilation for example - and can be related to guidelines for example those provided by CIBSE [9,10].
Box F is a measure of the efficiency of the equipment for which published benchmarks are available [9,11].
Box G is the number of hours per year that the service is required for each system.
Box H is a management factor which indicates the fraction of the time enabled the systems are working (for part load operation the hours are effectively reduced pro-rata).
To obtain these data for new building designs, the installed power densities (box C) and expected occupation hours (box G) are known, and management/control factors (box H) can be estimated from experience or published benchmarks [7]. Equivalent full load operating hours (box D) can therefore either be calculated or estimated directly from published values [8].
For occupied buildings, the installed power densities (box C) can be assessed on-site without excessive effort, and operating hours (box G) and control factors (box H) can be assessed by consideration of control schedules, observation and discussion with operators and occupants. This is the approach described in more detail in [12], used as the energy part of the Probe studies [13], and currently subject of development for publication by CIBSE.
This 4,000 mē headquarters building with both deep plan and narrow plan office areas has displacement ventilation and chilled beam cooling. The purpose of the energy assessment was to derive indices of annual consumption per mē of treated office floor area for the various building systems so that the building and each system's performance could be assessed in terms of appropriate yardsticks, and to explain these in terms of standards of service provided, appropriateness of installed plant and control.
Gas heating consumption (accounting for only 20% of the building's energy cost and energy-related CO2 emissions) is approximately equal to Typical as defined for prestige office buildings in the EEO office energy consumption guide [3]. This relatively high usage is mainly due to a leaky building envelope.
The building's electricity consumption (see Figure 2) appears at first sight to be between the Good Practice and Typical benchmarks. However, the "process" energy uses for this building (computer room, catering and office equipment) are all lower than in the benchmarks. If these consumptions are removed from the comparison, the remaining total building services energy use (up to and including the Other uses) is 15% higher than the Typical benchmark which leaves considerable room for improvement for a low energy design.
The Other energy usage includes 38 kWh/mē of humidification. Humidification energy is not included in the benchmarks which were developed when humidification was less common. Without humidification in this building, the building services energy would be 3% less than Typical - although still more than 50% higher than Good Practice.
The fans, pumps and controls usage is 25% more than Typical and more detail was needed for its analysis using information on separate fan and pump benchmarks [8]. Controls usage is quite low but the annual fan energy at 45 kWh/mē is between Good Practice and Typical when for this type of system low fan energy use should be achievable - the high consumption is caused by long running hours, and the system did not have variable speed (or two speed) drives to benefit from any reduced air volumes. The pump energy is extremely high because many pumps are on all year round and sometimes 24 hours (in one case with duty and standby pumps operating). The potential for saving in pump energy is clear.
Lighting energy use at 63 kWh/mē, although better than Typical, is higher than Good Practice despite a nominally efficient lighting system and sophisticated controls. The reason is that neither the power density of 18W/mē nor the full load operating hours (3500 hours) are as low as might be achieved: the level of lighting service at 500 lux has not been designed down to 400 or 350 lux as in some current buildings, the specific lighting efficiency is only moderate at 3.6 W/mē per 100 lux because of the light fittings and the use of uplighters for effect, and the controls/management factor is near to unity because the controls were ineffective: daylight dimming and occupancy sensing settings were set conservatively so that they rarely had effect, and the phased switching and maintained illuminance had not been implemented.
To achieve this analysis, the overall energy consumptions were obtained from utility bills. The remaining energy information was gathered over two visits by two people, noting equipment and plant power loadings, plant schedules and observing use of lighting and equipment. Further measurements included:
* Observation of main computer power demand as a standard menu option from the uninterruptible power supply operating panel. In this case it was also possible to clip-meter this supply.
* Clip-metering (using true power meters) of major fans, humidifiers, the UPS and computer room air conditioners.
* Portable demand profile metering of the two chillers for two minute periods over one hour.
* Daily meter readings by staff over three weeks in summer.
The clip metering and portable profile metering was completed within one day. It confirmed the crucial computer system energy use and associated air conditioning (being recirculating this will not be strongly weather dependent). It provided valuable support for equipment ratings obtained from nameplates and manuals, and indicated how the chillers were operating on a warm summer day from which seasonal operation could be deduced.
A set of three all-electric high street retail branches with extensive sales lighting and heat pump air conditioning were investigated to assess the overall performance against benchmarks, to show where energy is being used and to highlight energy wastage. The stores were selling dry consumer goods, with no product refrigeration, on one to three floors and ranged from 200mē to 800mē in sales floor area.
Electricity use per mē of sales floor area was compared with energy consumption yardsticks for all-electric non-food shops from the DoE's Introduction to Energy Efficiency in Shops and Stores [14], see Figure 3.
The electricity consumptions were generally high, attributed to the extensive sales area lighting and comfort cooling. While these yardsticks provide a useful base assessment, they may not adequately represent these buildings. The stores in this estate have fairly uniform sales activity, construction and servicing, so there is a case for developing customised yardsticks based on the results presented above; a more strategic review of lighting and servicing policy is also justified to ensure that this policy does not lead to excessive energy consumption generally for the estate.
High energy use by sales area lighting in store S3 is indicated so, given that the hours of use were typical, further analysis was carried out as summarised in Table 1. This showed that the perimeter sales lighting has extremely high installed power density at more than 300 W/mē, and that this is not purely due to high levels of illuminance because the specific consumption of 35 W/mē per 100 lux is a factor of nearly three above the value for store S1, and a factor of ten above general levels for fluorescent lighting common in offices.
Simple calculations based on Table 1 show that if the store S1 levels of perimeter lighting and efficiencies were used at store S3, the perimeter lighting demand would fall by 17 kW, a saving of 64% of the sales area lighting power demand. Whereas it was fairly clear that the perimeter lighting was inefficient, the extent of the inefficiency and the potential for saving were clarified by this analysis.
The relatively low air conditioning energy use at store S3 was attributed to its exposed location and good ventilation - so comfort cooling loads were not as high as might be expected with the high lighting gains, yet in winter some requirement for heating was offset by the lighting energy.
The higher "other" energy use in store S2 was due to computer and related equipment for an upgraded stock management system. The extra energy consumption and cost of this system was noted by the store management as part of the system trials.
To achieve this analysis, electricity consumptions were taken from annual electricity bills. Half hourly demand profiles were measured for a week to establish daytime, night-time and weekend energy usage. Equipment loadings were noted from rating plates and checked with clip ammeter and true power meter readings for desk top equipment and specialist equipment. Usage was established by inspecting timeswitches, by observation and discussion with management and staff.
The use of heat pumps for heating and cooling was assessed by inspection of the demand profiles in relation to the weather, and observation of plant operation over two visits during different seasons. Subsequently, detailed monitoring of other high street premises has been carried out to improve estimates of separate heating and cooling use.
The method for dealing with perimeter lighting areas and illumination levels was developed for this exercise and would need to be standardised for cross comparison with more widely differing types of premises.
The approach described provides a method for understanding and explaining the energy performance of a building and its systems. An investigation of total energy use (for a building or even for an end-use category) can only identify high consumption. The technique of considering all components in Figure 1 as measures of performance, with associated benchmarks, provides more valuable and practical information which can:
* improve the effectiveness of energy surveys by pinpointing problem areas, so that energy efficiency measures can implemented;
* provide a growing body of knowledge about how specific systems perform in different building types and environments, leading to practical detailed benchmarks;
* allow greater use of benchmarks in the briefing, design and operation of buildings.
A headquarters office example has shown how the approach can separate non-building services loads and explain the building services loads in terms of key systems performance.
An example for a set of three retail premises has shown how the approach can be used when detailed yardsticks are not available for system energy consumptions, and has indicated a poor performance and significant potential for saving from a sales area lighting system.
1 Introduction to energy efficiency in offices. Department of the Environment, 1994. (Also other guides in this series).
2 Practical energy saving guide for smaller businesses. Advisory committee on business and the environment/Department of Environment.
3 Energy efficiency in commercial and public sector offices. Energy Consumption Guide 19. BRECSU 1996. (Being revised, see reference 6 below).
4 Energy efficiency in industrial buildings and sites. Energy Consumption Guide 18. BRECSU 1993.
5 Energy efficiency in hotels - a guide for owners and managers. Energy Consumption Guide 19. BRECSU 1993.
6 Energy efficiency in offices. Draft revision of Energy Consumption Guide 19. BRECSU. Publication planned for late 1997.
7 Energy efficiency of buildings: Simple appraisal method. A Birtles and P Grigg. Building Services Engineering Research and Technology, Vol. 18 No 2 1997, pages109-114.
8 Technical review of office case study and related information. WT Bordass. General Information Report GIR15. Department of the Environment 1994.
9 Code for interior lighting. Chartered Institution of Building Services Engineers, 1994.
10 Environmental criteria for design. Guide Volume A1. Chartered Institution of Building Services Engineers, 1986.
11 Energy efficient design of new buildings except low-rise residential buildings (section 9.5.4). ASHRAE Standard 90.1, 1989.
12 Energy performance of occupied non-domestic buildings: Assessment by analysing end-use energy consumptions. J Field, J Soper, P Jones, W Bordass and P Grigg. Building Services Engineering Research and Technology, Vol. 18 No 1 1997 pp39-46.
13 Probe technical review. W Bordass, R Cohen and M Standeven. Building Services Journal April 1997, pp37-41.
14 Introduction to energy efficiency in shops and stores. Department of the Environment, 1994.