Case Study

Building Serv. Eng. Res. Technol.

2024, Vol. 0(0) 1–13
CIBSE TM54 energy projections II: A © The Author(s) 2024

case study using dynamic simulation Article reuse guidelines:

sagepub.com/journals-permissions
with template systems modelling DOI: 10.1177/01436244241245522
journals.sagepub.com/home/bse

Nishesh Jain1 , Esfand Burman1 and Dejan Mumovic1

Abstract
CIBSE TM54 was recently revised and covers best practice methods to evaluate the operational energy use of
buildings. TM54 is a guidance document on performance evaluation at every stage of the design and
construction process, and during the occupied stage, to ensure that long-term operational performance is in
line with the design intent. The main performance evaluation principles in TM54 are a step-by-step modelling
approach and scenario testing, to improve the robustness of the design proposal calculations. The latest
version brings an updated perspective to the modelling approaches, including dynamic simulation with
Heating, Ventilation and Air Conditioning (HVAC) systems. It also incorporates more detailed guidance
around risks, target setting, scenario testing and sensitivity analysis. A case study approach is used to explore
some of the important aspects described in TM54. TM54 recommends three modelling approaches (aka
implementation routes) that a project can follow depending on its scale and complexity: using quasi-steady
state tools; using dynamic simulation with template HVAC system; and using dynamic simulation with detailed
HVAC system modelling. As part of a series of three, this case study provides an application of the second
implementation route: using dynamic simulation with template HVAC.
Practical Application: This case study provides detailed guidance on undertaking CIBSE TM54 modelling
and projecting design stage building performance. The study covers the interpretation and clarifications of
how TM54 can be applied, through the dynamic modelling tools using template HVAC systems.

Keywords
CIBSE TM54, energy performance, performance modelling, dynamic simulation , simple HVAC

Received 29 December 2023; Revised 3 March 2024; Accepted 14 March 2024

Introduction 1
UCL Institute for Environmental Design and Engineering,
The Chartered Institution of Building Services En- University College London, London, UK
gineers (CIBSE) Technical Memorandum (TM)
54 provides building designers and owners with Corresponding author:
Nishesh Jain, UCL Institute for Environmental Design and
guidance on evaluating operational energy use once a Engineering, University College London, Gower St,
building’s design has been developed. First pub- Bloomsbury, London WC1E 6BT, UK.
lished in 2013,1 this was one of the first pieces of Email: [email protected]
2 Building Services Engineering Research & Technology 0(0)

industry guidance documents in the UK to address in this case study is based on the dynamic simu-
the performance gap issue and to better project op- lation method (DSM) with a template HVAC
erational performance of actual energy use. In the system and its subsequent steps are shown in
recently published revision to CIBSE TM54: 2022,2 Figure 1. The 17-step modelling methodology has
the guidance has been made up to date by taking been divided into three stages: baseline model
account of regulatory and industry changes such as generation, scenario/sensitivity assessments, and
net-zero transition, performance targets and advances result reporting & benchmarking. In the subse-
in Heating, Ventilation and Air Conditioning quent sections of the case study, the modelling is
(HVAC) modelling. presented following the steps presented in
For a holistic application and wider adoption of Figure 1, explaining the use of building-specific
best practice energy projections in the industry, the information in creating a TM54 model.
new TM54 document suggests three implementation Modelling for TM54 in this case is undertaken
routes for different scales and complexity of projects. using DesignBuilder Software,3 which is a graphical
The modelling approaches include: user interface for EnergyPlus.4 However, the input
definitions and explanations for building details,
1. Quasi-steady state modelling systems and the project context remain software
2. Dynamic simulation modelling (DSM) using agnostic. Similarly, results reporting along with
template HVAC systems process documentation of inclusion or exclusion of
3. Dynamic simulation modelling (DSM) using any aspects in building modelling also follow
detailed HVAC systems TM54 requirements. This makes the case study
presentation software agnostic and replicable for any
In this, second of three case studies, an example project that uses DSM with a template HVAC
of TM54 design stage building performance modelling approach.
modelling is presented for a school building using
dynamic simulation with template HVAC. First,
the step-by-step modelling methodology proposed Case study application and
in CIBSE TM54 for this implementation route is
described. Then the case study building is intro-
modelling approach
duced and the modelling inputs and assumptions The case study building is a secondary school and
for each TM54 step are described. Finally, results sixth form with academy status, located in London,
are presented per TM54 requirements including England. As part of a redevelopment project, six new
deterministic calculations, sensitivity and scenario buildings have been created for the school, and a
assessments, and benchmarking against industry couple of existing ones are retained. The buildings
standards. are four stories high, with a total useful floor area of
21,405 m2. In this case study, we focus on one of the
new teaching buildings (∼5000 m2), with typical
Methodology
educational activities such as classrooms, science
The main performance evaluation principles in labs, and faculty rooms. The project is planned to
TM54 are a step-by-step modelling approach, sys- have a biomass boiler using wood pellets and solar
tematic sensitivity and scenario analysis, and per- thermal collectors to meet the local council’s plan-
formance reporting including benchmarking, to ning conditions of having on-site renewable energy
improve the robustness of the design proposal cal- technologies.
culations and provision of advice to clients. This school is a large building with complex and
After Step 0, that is, the selection of the ap- varying operation and occupancy patterns, which
propriate modelling approach, depending on can be accurately modelled only by using a DSM
project-specific requirements, step-by-step mod- modelling approach. The HVAC system in the
elling is undertaken. The modelling approach used school building, has typical components and
Jain et al. 3

Figure 1. Step-by-step TM54 methodology for dynamic simulation using template HVAC.

simple controls. This means that a high level of modelling and forms the basis from which all fu-
detail in HVAC modelling and component-level ture calculations are undertaken. The process in-
energy breakdown is not necessary. Therefore, volves creating geometry, defining constructions,
DSM with a template-based HVAC modelling zoning, and selecting appropriate weather data.
approach is sufficient for this case study. Detailed The accuracy and quality of these inputs in the
guidance on modelling approaches and tools is model will in turn affect the accuracy of future
explained in TM54.2 Figures 2 and 3 show the model outputs.
building’s exterior image and a typical floor plan For this case study, the building’s architectural
respectively. design drawings (Figure 3) are used to inform the
inputs for geometry, construction, and zoning.
Construction details are defined with material layers
TM54 baseline and typical year-hourly weather data from the nearest
The baseline building generation is covered in weather station is selected for the simulations.
TM54 Steps 1 – 11. For each of the steps, model Figure 4 shows the model image and internal zoning
inputs must be defined along with ascertaining of the floor plans. In the model, various space types,
confidence levels in their assumptions to guide separated into zones, are teaching areas (60%), cir-
5.0 Scenario/Sensitivity analysis. culation (25%), offices for staff (6%), storerooms
(3%), high ICT rooms (3%), toilets (2%) and a server
room (1%).
Geometry and site details The external envelope is made of prefabricated
concrete panels, assembled at the site. The building,
Step 1: constructing and building the model
designed for high energy efficiency, has low fabric
Generating the model geometry and setting up site U-values and emphasises avoiding thermal bridging.
information is the first step in undertaking building Fabric properties are as follows: U-values (W/m2 °K)
4 Building Services Engineering Research & Technology 0(0)

Figure 2. Case study school building.

Figure 3. Typical floor plan.

for Wall: 0.25; Window: 1.6; Roof: 0.20; Ground: variation in the designed thermal properties will be
0.15; and design airtightness was 5 m³/hr/m2 @ lower than normal.
50 Pa. Glazing has a g-value of 0.26 and visible light
transmission (VLT) of 0.49. Appropriate insulation
has been added and glazing selection undertaken to
Operation details and internal gains
meet these values.1 Spaces are designed to have large Building occupancy numbers and other loads related to
windows for daylighting which are also partially internal gains such as lighting and equipment along
operable for natural ventilation and free cooling in with the intended hours of operation of the plant and
summer. equipment must be established as far as possible. They
The confidence level in the construction inputs is are typically estimated by assessing the installed
high due to the use of a prefabricated construction equipment and establishing the operating hours of the
process. Therefore, it is assumed that typical building and the way that the building is to be managed.
Jain et al. 5

Figure 4. Model visualisation of the school building and its internal zoning.

While design documentation can provide the nominal database.5 The other major space type covers office
occupancy and internal loads, information gathering rooms for teaching staff which are occupied by 4–
from the intended occupiers (for example by con- 5 staff resulting in an occupancy density of 0.10 people/
ducting a structured interview) is the best way to es- m2. For these three major space types, the occupancy
tablish building operation. In scenarios where this schedules used are shown in Figure 5.
information is not readily available, typical building
estimates can be used for the baseline modelling with a
high degree of uncertainty for scenario and sensitivity Step 3: lighting
analysis. Lighting gains are based on planned lighting
Data in the case study model about the occupancy equipment. The building is planned to have low-
and various internal gains are as per the design brief energy lighting (T5 fluorescent lamps) with an ef-
and from discussions with the client for loads and ficacy of at least 80 lm/W for teaching areas and
operation. Table 1 summarises the operation details offices and 65 lm/W for toilets and circulation areas.
and internal gains for this case study. The rest of the Different space types have target lighting levels set as
section explains these model inputs in detail, cov- 300–500 lux for teaching areas and 150–200 lux for
ering each of the steps given in TM54. circulation areas. These are combined with Passive
infrared (PIR) and daylight sensors. Lighting power
density is set to be 9 W/m2 as the building level
Step 2: estimating operating hours and average and its operation follows the occupancy
occupancy factors schedule for each of the spaces.
Teaching areas, the most common space type, have
about 35 persons in the classrooms. The average oc-
cupancy density of these areas is therefore 0.42 people/
Step 4: lifts and escalators
m2. Circulation corridors are used between classes and In this building, there is only one lift and therefore to
are sparsely occupied. Circulation occupancy density is simplify its calculations, it is modelled as small
assumed to be 0.11 people/m2 as per UK NCM power energy use in circulation areas.
6 Building Services Engineering Research & Technology 0(0)

Table 1. Operation details and internal gains for the case study building.

Operation
details Description Building detail Remarks

Step 2: This covers gains due to people The whole school campus is Model inputs (density and
Occupancy in the spaces defined by designed for 2000 pupils and schedules) for occupancy are
occupancy density and 250 staff. For building set zone-by-zone. This
schedules. Indoor simulation, varied occupancy enables the simulation of
environmental quality and and schedules are used for dynamically varying
operating hours for other different space types. occupancy-related loads and
loads are often linked to Teaching areas have an average any ventilation control
occupancy. They are density of 0.42 people/m2 with requirements.
described in their respective a varying schedule as per the
sections. school timetable. Overall
occupancy hours for weekdays:
08:30 – 15:00; sun & holidays:
nil
Step 3: Lighting Lighting energy use and Power density for classrooms, Total energy use and associated
associated heat gains are labs and offices is 9 W/m2 with heat gains for all these types of
estimated after defining its operation schedule as per the loads are estimated after
loads, operating hours, and occupancy. The building has defining their power ratings
controls. T5 fluorescent lamps, passive and operating hours.
infra-red (PIR) and daylight As DSM modelling is granular,
sensors. inputs for these loads are
Step 5: Small This includes plug loads, which Power density for equipment defined zone-by-zone with
power can be difficult to estimate, varies between 10 and 50 W/ hourly varying time schedules.
especially when the ultimate m2 across major spaces. The This enables the estimation of
occupier is not yet known. operation schedule is as per the their dynamically varying
Benchmark data can provide occupancy. energy use and their effect on
helpful estimates at the space-level heat balance.
design stage.
Step 4: Lifts and Design stage energy projections This building has a small
escalators must account for all end uses. proportion of load that could
Step 6: Catering Buildings may have some or be attributed to these
Step 7: Server all of these energy end-uses. additional loads. These included
Step 8: Plant and They have to be added lifts, IT and lab equipment.
equipment accordingly. These are added to zone-level
small power.
Step 9: Domestic Evaluating the energy used to This usage varies based on space Building services in the template
hot water provide DHW requires type and at the design stage, HVAC system are modelled
(DHW) estimating the amount of typical benchmarks are used, using ideal loads. DHW mains
water used, schedules, such as teaching areas’ demand and supply temperatures are
system energy losses, and that is set at 5 L/person/day. set as 10°C and 65°C.
fuel sources. DHW does not
affect the heat balance as
such.
a
This is a summary of model data used, in the implementation matrix; more detailed data needs to be recorded covering all space types so
that the modelling process can be replicated at later stages if needed.
Jain et al. 7

Figure 5. Occupancy schedules for the school building’s typical spaces.

Step 5: small power Step 9: domestic hot water (DHW)

The key plug-in energy end-uses in the building, The use of hot water is predominantly in the toilets. The
covered in small power, include computers, office typical consumption of hot water is estimated as 5 L/
equipment, lifts, and other plug-in loads. Small pupil/day as per CIBSE Guide G6 reference value for
power use is assumed to follow the same schedule as schools.
occupancy with 5% standby loads. Total loads are
defined as follows: Step 10: adding internal heat gains to the
model, for energy uses calculated outside
· 1100 W per Teaching Area (or 13.75 W/m2
the model
based on 80 m2 classroom/lab)
· 50 W/m2 in ICT-rich areas Table 1 covers all the model data inputs from operations
· 10 W/m2 for office areas and gains as per TM54 Steps 2 to Step 9. Following this,
· 2-10 W/m2 in circulation and toilets Step 10 requires adding internal heat gains to the model,
for energy uses calculated outside the model. As the
Step 6: catering modelling process is undertaken using a DSM tool, the
impact of internal gains on the heat balance is already
This load is not represented in this building, as there included in the calculations. There were no additional
is a separate building for catering and dining in this energy uses that were calculated outside the model.
school. The confidence level in the occupancy-related
factors (Steps 2–9) is low, similar to most buildings
Step 7: server at the design stages. Therefore, a typical variation is
expected to be seen in these inputs from what is es-
The building server room’s planned capacity is set to timated here and in the actual building. These variations
be 25 kW which equates to 1000 W/m2. Its operation can typically be up to ±10% as per BS EN 15603:20087
is set to be at full capacity during occupied hours and and have been used to inform the sensitivity analysis.
at 10% standby capacity during non-occupied hours.
HVAC system definition
Step 8: plant and equipment
Step 11: modelling HVAC systems and
The building does not have any significant additional
their controls
equipment. The security system and any other plug
loads are already added in step 5: small power The energy use for HVAC systems and associated
section. controls are required to be estimated and separately
8 Building Services Engineering Research & Technology 0(0)

reported. This energy use should include all HVAC Building services operations are linked to occu-
systems including heating and cooling, fans and pancy patterns. Space conditioning systems are
pumps, and domestic hot water (unless calculated planned to be turned on 2 h before classrooms are to
outside the model). be occupied until the end of the classes. The heating
In Step 11, template HVAC modelling is selected, setpoint is set to be 20°C and in spaces where cooling
and key model inputs are customised such as is provided, the cooling setpoint is set to 23°C. The
equipment coefficient of performance (CoP), mechanical ventilation rate for the key occupied
whereas other more detailed inputs such as part load spaces varies from 5 to 12 L/s/person with specific
performance and controls configuration are assumed fan power calculated as 1.8 W/l/s.
to have typical behaviour. Modelling of this system in this implementation
In this case study, a simple HVAC system is route is carried out by using the Simple HVAC
proposed. Heating is to be provided through a modelling route in DesignBuilder. In ‘Simple HVAC’
centralized plant for the entire campus via a the heating/cooling system is simulated using the
pressurised low-temperature hot water (LTHW) basic EnergyPlus ZoneHVAC:IdealLoadsAirSystem
system. A biomass boiler (heating seasonal effi- method.8 This supplies hot/cold air to meet heating
ciency: 0.75) will provide heat for annual DHW and cooling loads. Mechanical ventilation loads are
demand and two gas-fired boilers (heating seasonal also calculated locally for each zone. Subsequently,
efficiency: 0.84) will provide the remaining heat fuel energy consumption for the boiler and VRF
and function as a backup to the biomass boiler. system is calculated from zone heating and cooling
ICT-enhanced spaces will have Variable Refrig- loads as a post-process calculation using simple
erant Flow (VRF) systems that provide both seasonal efficiency factors.
heating and cooling in labs and cooling in the
server room. (heating/cooling seasonal efficiency:
1.47/3.80). A mechanical ventilation system with
TM54 baseline results
heat recovery (efficiency: 0.75) via a centralised As the key modelling inputs have been set up by Step
roof-mounted AHU will provide fresh air in the 11, a simulation for the TM54 baseline performance
building, distributed through wall-mounted projection can be run. The building’s energy use
diffusers/grills. BMS-control, based on CO2 sen- projection is 161 kWh/m2/annum. Figure 6 shows
sors will also be used to provide the appropriate the breakdown of projected energy uses, separated by
amount of fresh air. end-use categories.

Figure 6. TM54 baseline energy use per end use for the baseline building.
Jain et al. 9

Heating is supplied by the central (biomass and those parameters which are least certain and/or are most
gas) boilers at the facility level and the building level influential. This allows quantification of the difference
the reported energy use is for this building only, between nominal performance (Figure 6, at the end of
presented as energy use based on the central system’s Step 11) and how the building is likely to operate. This
CoP. Space heating, hot water and CO2-based de- risk assessment can be done using a systematic ap-
mand control ventilation systems (auxiliary energy) proach of sensitivity (Step 13) and scenario analysis
account for a major proportion of energy use for the (Step 14) which can identify the most important and
building. This is the central estimate for energy influential model inputs and quantify the total vari-
performance. However, for better communication of ability in the calculation results.
expected building performance, realistic scenarios As a part of the TM54 baseline modelling process
that account for uncertainty in modelling assump- (Steps 1–11), all key input parameters were tagged with
tions should also be presented. the confidence level in their values. For this building,
the impact of key model input parameters which are
uncertain such as occupancy and internal gains or
Scenario/sensitivity analysis which can significantly influence building performance
Performance projections at the design stage are prone to such as construction quality or weather data are ex-
several operational risks. These risks can be due to plored in the sensitivity analysis. Table 2 lists all the
management of the building after occupancy (Step 12) uncertain variable categories, the model inputs that are
or functional changes that occur in the building over changed, and their variation (lower limit, upper limit,
time. To account for these risks, when projecting energy and worst case) assumed in the sensitivity analysis. The
use, a range of simulation runs should be undertaken. variations are those that are assumed by modellers
These runs should consider the variety of plausible based on their experience, published literature,7,9 and
potential real-world operating scenarios, focusing on discussions with key stakeholders and building users.

Table 2. Design variables explored in the sensitivity analysis.

Variable Lower Upper Worst

category Model input limit limit case Remarks

Geometry Wall U-values Val 5% Val +5% Val +10% Despite the use of prefabricated panels, insulation
Window Val 5% Val +5% Val +10% was applied on site, therefore variation due to
U-value construction quality issues is possible.
Roof U-value Val 5% Val +5% Val +10%
Occupancy Occupancy Val 10% Val +10% Val +20% This is a typical variation seen in occupant-related
density behaviour. Confidence is low in these variables
Internal gains Equipment Val 10% Val +10% Val +20% because typical school usage evolves.
power density
Lighting power Val 10% Val +10% Val +20%
density
DHW DHW demand Val 20% Val +20% Val +25%
HVAC Heating setpoint Val 1°C Val +1°C Val +2°C
Cooling setpoint Val +1°C Val 1°C Val 2°C
Boilers Val Val 5% Val 10% Variation is possible due to commissioning and
efficiency operation issues. The confidence level is
VRF efficiency Val Val 5% Val 10% moderate as the HVAC strategy is simple.
Specific fan Val 10% Val +10% Val +20%
power

Lower limit: value that will lead to lower energy use; Upper limit: value that will lead to higher energy use; Worst case: value that will lead
to even higher energy use due to significant management and operation issues.
10 Building Services Engineering Research & Technology 0(0)

Parametric simulations run by changing multiple pa- building. Future climate scenario test is undertaken
rameters are undertaken to ascertain the most important using 2050 weather data for high, medium, and low
design inputs that cause variation in building perfor- emission scenarios2 and the total energy use pro-
mance from the central (TM54 baseline) estimate. jection is 93 kWh/m2/annum, 102 kWh/m2/annum
Regression-based parametric sensitivity anal- and 114 kWh/m2/annum respectively. It should be
ysis was undertaken by running 250 simulations noted that with increasing temperatures the building,
and changing multiple variables at a time. This not having comfort cooling for most of the spaces,
analysis assumes that design variables are inde- may experience overheating in summer and this may
pendent of each other. Figure 7 shows the sensi- necessitate the installation of a cooling system. Such
tivity analysis results for total building energy use, an analysis can inform design decisions and rec-
ranking the parameters from highest to lowest ommendations for the long-term management of the
importance. school.
The variables with the highest standardised re- Besides the upper and lower range of likely building
gression coefficient (SRC) are the most important. For performance, the worst-case scenario energy use is also
example, within the range of variations assumed, the calculated as 194 kWh/m2/annum. This worst-case
most influential parameters are heating setpoint, system assumes a poorly managed building with the values
efficiency, and fabric u-values. All these parameters under the worst-case column in Table 2 such as ex-
affect the largest energy end-use (heating energy) tended running hours, high occupancy, and high levels
significantly and therefore rank highly in the impact on of internal loads. The current worst case is significantly
total energy use. The direction of the SRC shows a higher than the central energy use projection and
direct or inverse relationship. For example, if the highlights the significance of managing the perfor-
heating setpoint increases, the total energy use will mance in use.
increase, however, if boiler efficiency increases the
overall energy use decreases. Besides these DHW
demand is also a highly influential design variable on
Reporting and benchmarking
total energy use. Presenting simulation results in context and com-
Uncertainty is also calculated by varying all the paring them against benchmarks and building targets
input parameters in Table 2 within the ranges defined, can be useful to determine whether the results are
the same set of simulations used in the sensitivity within an acceptable range. Figure 9 shows the
analysis above. Figure 8 shows the impact of vari- comparison of TM54 baseline estimates against the
ability (uncertainty) of all these inputs on individual good practice (25th percentile) and typical (median)
end-uses and the total energy use. The total energy benchmarks as per the DEC database,10,11 and
use of this case study building can range between CIBSE TM46 benchmark.12
113 kWh/m2/annum and 180 kWh/m2/annum. It is seen that the TM54 central estimate is below
Therefore, if the building is to perform as intended typical benchmarks. However, the uncertainty (seen
the most influential parameters discussed above in Figure 8) and the worst-case performance pro-
would require close monitoring and safeguarding jection mean that it is possible for the building to
from significant changes. exceed the performance benchmarks if safeguards
Besides the sensitivity and uncertainty analysis, are not put in place to manage energy use during
two distinct scenario analyses were also undertaken: operation.
Result reporting for TM54 calculations should not
1. Future climate scenario, just be limited to central estimates, as seen in
2. Worst-case scenario. Figure 6, but also include further results and as-
sessments undertaken after uncertainty and sensi-
For future climate scenarios assessment, the tivity analysis along with a comparison against
ambient temperatures are expected to increase, benchmarks. Reporting of modelling inputs and
leading to a reduction in the heating energy use of the outputs in a structured implementation matrix
Jain et al. 11

Figure 7. Sensitivity analysis of variable inputs on total energy use.

Figure 8. Uncertainty around TM54 baseline energy use for various end uses.

(Annex A of CIBSE TM542) allows for documen- Operational performance of

tation of all the assumptions that have been made the building
alongside the results, at each modelling step. While
this keeps the modelling process transparent, it also This case study is an example application for CIBSE
provides useful documentation for operational stage TM54, which focuses on the design stage projection
performance assessments. of energy use. These design calculations are based on
12 Building Services Engineering Research & Technology 0(0)

Figure 9. Simulated energy use compared to industry benchmarks.

the best estimates that are available to the modelling for modelling and diagnosis during post-occupancy
team during the design stage. The actual energy use evaluation.
can be very different, even when calculated with best
practice modelling at the design stage. Using
TM54 performance modelling the compliance gap2
Summary
can be eliminated but a performance gap may still The following are the key points emerging from the
occur due to many changes that might happen in a CIBSE TM54 energy projections for this case study
building including those that cannot be estimated at school building using dynamic simulation with
the design stage. The underlying causes of an energy template HVAC:
performance gap go beyond the scope of modelling
in TM54 and its accuracy. These causes can be · Performance assessment of this school, a large
mapped to several factors across various construction building with varying operational patterns but
stages.13 a rather simple HVAC strategy, was under-
The building’s energy use is metered as taken using dynamic simulation modelling and
230 kWh/m2/annum. This is higher than the central a template-based HVAC system approach.
estimate in ‘TM54 baseline results’ section and also · Step-by-step modelling as per TM54, factoring
outside the uncertainty range calculated in ‘sce- in all energy end-uses and operational details,
nario/sensitvity analysis’ section. The identification resulted in an energy use projection of
of the root causes of the deviation is beyond the 161 kWh/m2/annum.
scope of TM54 methodology and this case study · Within reasonable uncertainty of model inputs
document. These causes have been determined and operational changes, the total energy use
using CIBSE TM63 methodology14 and described of this case study building can range between
in a separate paper.15 CIBSE TM63 Operational 141 kWh/m2/annum and 180 kWh/m2/annum.
performance: Modelling for evaluation of energy in However, in case of severe mismanagement
use14 provides a calibration-based modelling and operational issues, this can increase to
framework and a step-by-step guide for measure- 212 kWh/m2/annum.
ment and verification of energy performance in use. · Sensitivity tests show that the building per-
The framework is designed to determine the energy formance is highly susceptible to under-
performance gap in operation concerning design performance due to climate change and
calculations and identify the root causes of the uncertainties in building operation. This type
gap. It builds upon the design stage modelling done of assessment is necessary to have a better
as per TM54 and is the natural successor of this TM understanding of the operational risks and
Jain et al. 13

potential mitigation measures that could be 3. DesignBuillder Software Ltd. DesignBuilder V7.
considered at design stages and throughout the Stroud, UK: DesignBuillder Software Ltd., 2023.
life cycle of a building. 4. NREL. EnergyPlus V9.4. Golden, CO: National Re-
newable Energy Laboratory, 2021.
Acknowledgements 5. BRE. UK NaUK’s national calculation method for
The authors wish to express their gratitude to the designers, non domestic buildings. [Online]. Watford, UK:
building managers and users who engaged in research and BRE, https://www.uk-ncm.org.uk/ (2022, accessed
supported the building performance evaluation. We spe- 19 12 2023).
cially acknowledge the continued support of AHMM ar- 6. CIBSE, Public health engineering CIBSE guide G,
chitects, the company which was responsible for the London, UK: The Chartered Institution of Building
building’s architecture. Photographs in Figure 1 were taken Services Engineers, 2014.
by Tim Soar. 7. BSI. BS EN 15603:2008 energy performance of build-
ings. Overall energy use and definition of energy ratings.
Declaration of conflicting interests London, UK: British Standards Institution, 2008.
8. EnergyPlus Development Team. Ideal loads air system -
The author(s) declared no potential conflicts of interest
engineering reference — energyplus 9.4. [Online], https://
with respect to the research, authorship, and/or publication
bigladdersoftware.com/epx/docs/9-4/engineering-
of this article.
reference/ideal-loads-air-system.html#ideal-loads-air-
Funding system (2020, accessed 31 May 2023).
9. Macdonald I. Quantifying the effects of uncertainty in
The author(s) received no financial support for the
building simulation. PhD thesis, University of
research, authorship, and/or publication of this article.
Strathclyde, Department of Mechanical Engineering,
Glasgow, UK, 2002.
ORCID iD
10. Godoy-Shimizu D, Armitage P, Steemers K, et al.
Nishesh Jain  https://orcid.org/0000-0002-4116-0903 Using display energy certificates to quantify schools’
energy consumption. Build Res Inf 2011; 39(6):
Notes 535–552.
1. For full TM54 reporting the fabric details would be 11. CIBSE. Energy benchmarking dashboard. [Online].
explained in more detail including construction layers London, UK: CIBSE, 2023, https://www.cibse.org/
for different elements and thermal mass. knowledge-research/knowledge-resources/
2. Future weather scenario modelled was as per the Com- knowledge-toolbox/energy-benchmarking-dashboard
monwealth Scientific and Industrial Research (CSIRO) 12. CIBSE. TM46: energy benchmarks. London, UK:
future climate model for IPCC’s B1 storyline, downloaded CIBSE, 2008.
from the DesignBuilder Climate Analytics platform 13. CIBSE. Operational performance of buildings CIBSE
(https://designbuilder.co.uk/software/climate-analytics) TM61. London, UK: Chartered Institution of Building
Services Engineers, 2020.
References 14. CIBSE. Operational performance: building perfor-
1. CIBSE. TM54: evaluating operational energy per- mance modelling and calibration for evaluation of
formance of buildings at the design stage. London, energy in-use CIBSE TM63. London, UK: Chartered
UK: The Chartered Institution of Building Services Institution of Building Services Engineers, 2020.
Engineers (CIBSE), 2013. 15. Jain N, Burman E, Stamp S, et al. Cross-sectoral
2. CIBSE. TM54 evaluating operational energy use at assessment of the performance gap using calibrated
the design stage. London, UK: The Chartered Insti- building energy performance simulation. Energy
tution of Building Services Engineers (CIBSE), 2022. Build 2020; 224: 110271.