CIBSE TM54 energy projections I: A case study using quasi-steady state modelling

Introduction

The Chartered Institution of Building Services Engineers (CIBSE) Technical Memorandum TM54 provides building designers and owners with guidance on how to evaluate operational energy use once a building’s design has been developed. First published in 2013, ¹ this was one of the first pieces of industry guidance documents in the UK to address the performance gap issue and better project operational performance of actual energy use. In the recently published revision, CIBSE TM54: 2022, ² the guidance has been made up to date by taking account of regulatory and industry changes such as the net-zero carbon transition, performance targets and advances in Heating, Ventilation and Air Conditioning (HVAC) modelling.

For a holistic application and wider adoption of best practice energy projections in the industry, the new TM54 document suggests three implementation routes depending on different scales and complexity of projects. The modelling approaches include:

1. Quasi-steady state modelling

In this first of three case studies, an example of TM54 design stage building performance modelling for a Passivhaus school building using a steady-state modelling tool is presented. In the first part of this case study, the step-by-step modelling methodology proposed in CIBSE TM54 for the steady-state modelling implementation route is described. Then the case study building is introduced and the modelling inputs and assumptions for each of the TM54 steps are explained. Finally, results are presented as per TM54 requirements including deterministic calculations, sensitivity and scenario assessments, and benchmarking against industry standards.

Methodology

The main performance evaluation principles in TM54 include a step-by-step modelling approach, systematic sensitivity and scenario analysis, and performance reporting including benchmarking, to improve the robustness of the design proposal calculations and provision of advice to clients.

After the selection of the appropriate modelling approach (Step 0), depending on project-specific requirements, step-by-step modelling is undertaken. The modelling approach used in this case study is quasi-steady state modelling. The procedural steps involved as per TM54 are shown in Figure 1. The 17-step modelling methodology has been divided into three stages: baseline model generation, scenario/sensitivity assessments, and result reporting & benchmarking. In the later sections of the case study, the modelling is presented by following these steps, explaining the use of building-specific information in creating a TM54 model.OPEN IN VIEWER

Modelling for TM54 in this case study is undertaken using, the Passive House Planning Package (PHPP) tool. ³ PHPP is used for buildings aiming to achieve Passivhaus certification and represents quasi-steady tools used for the evaluation of energy performance in buildings. PHPP, when used for PassivHaus certification, may use certain standard model inputs. However, in this case study PHPP is used for TM54 modelling and therefore the inputs in the PHPP tool are building-specific. Also, the presentation of building details and the project context follow TM54 steps and may not be in the same workflow as the software itself. Similarly, the results reporting along with process documentation of inclusion or exclusion of any aspects of modelling also follow TM54 requirements. This makes the case study presentation software agnostic and replicable for any project that uses a quasi-steady state modelling approach.

Case study application and modelling approach

The school is approximately 1,500 m² and is located in Trimsaran, Carmarthenshire County in rural Wales. Being a primary school, the two-storey building mainly consists of Classrooms (40%), Circulation/Hub (30%) and halls (15%) with the remaining areas including storage rooms, offices for staff, toilets and kitchen. The school has been designed to meet the Passivhaus standard of energy efficiency and comfort. Passivhaus is an energy standard to achieve the lowest practical heating demand for a building . ⁴ The building meets the standard requirements by having high levels of insulation, triple-glazed windows, and mechanical ventilation with heat recovery (MVHR) among other features. Figure 2 shows the building’s exterior image.OPEN IN VIEWER

The compact design of this school building with highly insulated envelope and airtightness means that the heat gains/losses are minimised. Additionally, as the building uses simple heating and ventilation systems with predictable operational patterns, it can be assumed that the building energy demand will not vary significantly with hourly external weather. Therefore, it is suggested that a simplified approach to modelling, such as quasi-steady state building energy modelling can be used to develop the TM54 model of this building. Detailed guidance on modelling approaches and tools is explained in TM54. ²

TM54 baseline

The baseline building generation is covered in TM54 Steps 1 – 11 (Figure 1). For each of these steps, model inputs must be defined along with ascertaining confidence levels in the underlying assumptions to guide 5.0 Scenario/Sensitivity analysis.

Geometry and site details

Step 1: Constructing and building the model

Defining the building geometry and setting up the site information in the model is the first step in undertaking building modelling and forms the basis from which all future calculations are undertaken. This process involves setting up building geometry details and areas, defining constructions, zoning, and selecting appropriate weather data. The accuracy and quality of these inputs in the model will in turn affect the accuracy of future model outputs. Contrary to DSM tools, where building geometry and zoning are defined in three dimensions, spreadsheet-based tools such as PHPP require simpler inputs (calculated precisely from the actual drawings) such as overall zone areas and volumes and envelope component details.

For this case study, the building’s architectural design drawings, such as the typical floor plan presented in Figure 3, are used to inform PHPP inputs sheets, such as:

• Fabric U-values & airtightness: In this sheet construction assemblies for wall, floor and roof fabric elements are defined for use in heat transfer calculations. The building, designed for high energy efficiency, has low U‐values and emphasises avoiding thermal bridging. U‐values (W/m²K) are as follows: Wall: 0.144; Floor: 0.131; Roof: 0.109; and air permeability as 0.64 m³/hr/m² @ 50 Pa. ¹

• Areas: For building assembly (zones, and fabric elements), the total thermal envelope surface area is 3,129 m² and Treated Floor Area (TFA) is 1,327 m². Thermal bridges are defined along with areas in the same sheet in PHPP. The total length of thermal bridges to ambient air (psi: 0.165 W/mK) is 29 m and the length of the perimeter thermal bridge (psi: 0.017 W/mK) is 10 m.

• Ground: Details including soil properties (conductivity: 2.0 W/mK; Heat capacity: 2.0 MJ/m²K) and ground slab details (U‐value: 0.131 W/m²K; area: 374 m²; perimeter: 125 m) are defined for accurate calculation of ground heat transfers.

• Windows and doors: Geometry and frame details for windows (window area: 277 m²; glazing area: 196 m²), orientation (north: 20%; east: 5%; south: 65%; west: 10%) and thermal properties (glazing U-value: 0.6 W/m²K; glazing g-value: 0.5; frame U-value: 0.73 W/m²K; door U-value: 1.4 W/m²K) including thermal bridging of glazing (psi window: 0.04 W/mK; psi door: 0.1 W/mK).

• Shading: Details of shading from building elements (overhangs and reveals) to calculate the shading coefficient for each of the windows along with shading from surrounding buildings.

OPEN IN VIEWER

Figure 3.

First-floor plan of the school.

Climate data for Zone 14-Wales is used from the 22 climate data sets in the UK included within PHPP. This data is generated using ‘Met Office’ data and ratified by the Passivhaus Institute. ⁵ It contains details such as average monthly temperatures (external, dew point, sky, and ground) and radiation levels (horizontal and for each direction).

The confidence level in the construction inputs is considered high due to the building being designed to high Passivhaus standards. Therefore, it is assumed that typical variation in the designed thermal properties will be negligible.

Scenario/sensitivity analysis

Performance projections at the design stage are prone to several operational risks. These risks can be due to variations in the specification or performance of building systems under real operating conditions, management of the building (Step 12) or functional changes that occur in the building over time. To account for these risks, a range of simulation scenarios should be undertaken. These simulations should consider the variety of plausible potential real-world operating scenarios, focusing on those parameters which are least certain and/or are most influential. This allows quantification of the difference between the nominal (idealised) performance (Figure 4, at the end of Step 11) and how the building is likely to operate in practice. This risk assessment can be done using a systematic approach for sensitivity and scenario analysis which can identify the most important and influential model inputs (Step 13) and quantify the total variability in the calculation results (Step 14). As a part of the TM54 Baseline modelling process (Steps 1–11), all key input parameters were tagged with the confidence level in their values.

For this building, four different scenarios/sensitivity assessments were undertaken to quantify the impact of key uncertain model input parameters on the energy use outputs.

1. Future climate scenario assessment.

Table 1 lists all the uncertain variable categories, the model inputs that are changed, and their variation (lower limit, upper limit, and worst case) assumed in the sensitivity analysis. The variations are those that are assumed by modellers based on their experience, published literature,^6,7 and stakeholder discussions. Being a Passivhaus design, construction-related uncertainties are expected to be very low. Therefore, it was assumed that the effect of variation in these input parameters on energy performance results would be negligible.

For future climate scenarios assessment, the ambient temperatures are expected to increase, leading to a reduction in the heating energy use of the building. The weather data used ³ for the lower limit was 2050 high emissions compared to the 2050 low emissions scenario being used as the upper limit because in the former the increase in temperature is greater. The results showed only a slight decrease in total energy use in both cases because heating accounts for a lower proportion of total energy. For the lower limit and upper limit, the energy use is 39 kWh/m²/annum and 40 kWh/m²/annum respectively. It should be noted that with increasing temperatures, the building, not having comfort cooling, will see a significant amount of overheating in summer and this may necessitate the installation of a cooling system in the future. Such an analysis can inform design decisions and recommendations for the long-term management of the school.

Figure 5 shows the differential sensitivity analysis of total energy use in the building. Multiple calculations were undertaken making one change at a time for each of the model input categories listed in Table 1. The bar represents the central (nominal) energy use, and the whiskers show the high and low energy use for the variability of that particular input parameter. It can be observed that, within the range of variability assumed, occupancy and occupant behaviour (hours of use, lighting and power density, and heating setpoints) and mechanical ventilation are more influential parameters that affect the total energy use of the building. The occupant-related factors are significantly more important than the external environment due to the school having an extremely thermally efficient and airtight envelope.OPEN IN VIEWER

OPEN IN VIEWER

Table 1.

Design variables explored in the Scenario/Sensitivity analysis.

Variable category	Model input	Lower limit	Upper limit	Worst case	Remarks
Location	Climate data (future climate scenario analysis)	2050 high emissions	2050 Low emissions	-	Data changed to future climate values that represent different emission scenarios
Occupancy	Occupancy density	Val − 10%	Val + 10%	Val + 20%	This is a typical variation assumed for this case study in the occupancy-related behaviour. The confidence level is low in these because building usage evolves
	Hours of use	Val − 10%	Val + 10%	Val + 20%
Internal gains
	Lighting and power (incl. Lifts, servers)	Val − 10%	Val + 10%	Val + 20%
HVAC	Heating setpoint	Val − 1°C	Val + 1°C	Val + 2°C
	Mech vent rate	Val − 10%	Val + 10%	Val + 20%
	Boilers efficiency	Val	Val − 5%	Val − 10%	The lower limit is the same as the baseline and the upper limit is negative as the efficiency is likely to be lower if there are commissioning and operating issues. Being a simple system, the confidence level is moderate
	Heat recovery efficiency	Val	Val − 10%	Val − 20%

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.

Combining the variations for the input parameters in Table 1, the overall variation (uncertainty) in total energy use output ranges between 33 kWh/m²/annum for the lower limit and 52 kWh/m²/annum for the upper limit. Therefore, if the building is to perform as intended the most influential parameters listed above would require close monitoring and safeguarding from significant changes.

Besides the upper and lower range of likely building performance, the worst-case scenario energy use is also calculated as 68 kWh/m²/annum. This worst-case assumes a poorly managed building with the values under the worst-case column in Table 1.

Reporting and benchmarking

Presenting simulation results in context, comparing them with benchmarks, and building targets can be useful to determine whether the results are within an acceptable range. Figure 6 shows the comparison of TM54 baseline estimates against the good practice (25^th percentile) and typical (median) benchmarks as per the DEC database, ⁸ and CIBSE TM46 benchmarks. ⁹ It is seen that the TM54 central estimate is almost 70% below good practice benchmarks, especially due to very low heating (fossil fuel) energy use. For electricity, the central estimate is also better than the good practice benchmark, however, the uncertainty (seen in Figure 5) and the worst-case performance projection mean that it is possible for the building to exceed this value if safeguards are not put in place to manage electricity use during operations. In either case, however, the total energy use is expected to remain about 50% below benchmark aggregates.OPEN IN VIEWER

Result reporting for TM54 calculations should not just be limited to central estimates, as seen in Figure 4, but also include further assessments undertaken after scenario/sensitivity analysis along with a comparison against operational benchmarks. Reporting of modelling inputs and outputs in a structured implementation matrix (for an example, see Annex A of CIBSE TM54 ² ) allows for documentation of all the assumptions that have been made alongside the results, at each modelling step. While this keeps the modelling process transparent, it also provides useful documentation for operational stage performance assessments.

Operational performance of the building

This case study is an example application for CIBSE TM54, which focuses on the design stage projection of energy use. These design calculations are based on the best estimates that are available to the modelling team during the design stage. The actual energy use can be very different, even when calculated with best practice modelling at the design stage. Using TM54 performance modelling the compliance gap ² can be eliminated but a performance gap may still occur due to many changes that might happen in a building including those that cannot be estimated at the design stage. The underlying causes of an energy performance gap go beyond the scope of modelling in TM54 and its accuracy. These causes can be mapped to several factors across various construction stages. ¹⁰

The building’s energy use in the second year of operation is metered as 48 kWh/m²/annum. This is slightly higher than the central estimate in ‘TM54 baseline results’ section but within the uncertainty range calculated in ‘scenario/sensitvity analysis’ section. The identification of the root causes of the deviation is beyond the scope of TM54 methodology and this case study document. CIBSE TM63 Operational performance: Modelling for evaluation of energy in use ¹¹ provides a calibration-based modelling framework and a step-by-step guide for measurement and verification of energy performance in use. The framework is designed to determine the energy performance gap in operation concerning design calculations and identify the root causes of the gap. It builds upon the design stage modelling done as per TM54 and is the natural successor of this TM for modelling and diagnosis during post-occupancy evaluation.

Summary

The following key points emerge from the CIBSE TM54 energy projections for this case study school building using a quasi-steady-state modelling approach:

• The school building has a compact design and highly insulated envelope with low air permeability. Additionally, as the building uses rather simple heating and ventilation systems with predictable operational patterns, it can be assumed that building simulation results will not vary significantly with slight changes in heat balance or hourly external weather. These considerations can help determine which TM54 modelling route is the best option for a given context.