Skip to content Skip to navigation

Assessment and Reduction of the Risks Inherent in Performance Based Contracting by Use of CIFE’s Automated Building Energy Model Generator

Project Team

Catherine Gorlé , Martin Fischer, Sergio Tarantino, Iro Armeni
 

Overview

With ever more stringent building regulations and codes being introduced at Federal and State levels, including the mandated push towards NZEB buildings in Europe, the use of Building Energy Models (BEM) has become more commonplace. The accuracy of these models is still questionable, based on traditional manual, time-consuming methods of BEM generation.
In parallel, there is growing trend towards outsourcing of the energy management of commercial buildings. This outsourcing usually requires one party to contractually deliver a near constant level of occupant comfort while the other party pays a fixed annual fee for this service. The arrangement is known as Performance Based Contracting or PBC. To date, PBC has been fraught with unacceptable risks of performance and costs uncertainty due to misalignment between predictions and measured energy consumption, particularly on the side of the supplier, because a deep understanding of the building and plant operations has been difficult to assess prior to entering such contracts.
In order to assess and minimize these operational risks for all parties, researchers at CIFE are developing an automated rapid method to gather accurate building geometry, starting with a point-cloud from a laser scan, allows the BEM modeler to generate an accurate baseline for any building. Coupled with a parallel development in estimation of observed U-Values for building envelopes, an accurate BEPS model can now be created quickly and without building intrusion. To make this automated development useful, it is important to assess and quantify the parametric error associated with these BEPS models. Starting with Monte Carlo simulations, several statistical methods will be applied to assess the risk involved in using the BEPS as a foundation for Performance Based Contracting (PBC) on existing buildings.

Project Background

Research Motivation

The promise of using Building Information Model (BIM) techniques to help in the development of accurate Building Energy Performance Simulation (BEPS) models has not materialized, in particular for existing buildings. Much has been spoken about the promise but this comparatively informal approach is occurring while BIM is now mandated for all Government building work in the UK in a fully formal sense.
 
BEPS models are often developed for new buildings as a means of showing compliance with whatever code or regulation is required in that geographical location. In many cases, the models are discarded after that compliance purpose has been served (see Figure 1). The models’ accuracy is very seldom tested, post occupancy. When it is tested, it is usually done as part of an academic exercise and commonly shows the model to forecast about ±50% of the actual energy usage, post occupancy. Furthermore, the cost associated with model development is high. All of these factors have combined to give the impression of a lack of quantified accuracy and reliability, and ultimately, model usefulness.
 
For precisely these reasons, BEPS models are rarely developed for existing buildings, particularly due to their associated high cost. Older existing buildings present a special challenge given the frequent lack of availability of as-built drawings, and the manual data collection for internal geometry can prove too difficult, if not impossible.

Research Objectives

Recent research work involving CIFE and CS in Stanford has very effectively shown how data from a laser scanned building interior may be used to automatically generate a full and accurate BIM. This has been referred to as the Building Parser (BP).
 
A 2017 follow-on CIFE study involving building energy modelling (BEM), it has been found that by redefining and re-applying the rules by which the laser scan point cloud data are interpreted to generate an accurate BEPS model, i.e. the semantic definitions, the geometry specifically required by the energy modelling systems was automated generated without the BIM step.
 
The simplified BEPS study by CIFE has shown how low cost, accurate and rapidly developed BEPS models can be delivered. It is well accepted and this CIFE work has confirmed the need for accurate geometry. Simulation systems such as EnergyPlus or IES’ Virtual Environment (VE) heavily depend on this accuracy. 
 
This project is proposed as a contribution to two fields of interest at CIFE:
1. Confirm the Scan-to-BEM method to automatically generate accurate and usable geometry suitable for direct input to energy simulation systems
2. Include the U-Value estimation protocol to complete a full BEPS model for each building
3. Test the accuracy of the geometry and materials importation in generating a BEPS model for a commercial building by comparing to a known good conventionally generated model
4. Develop a statistical method to assess the accuracy required of the BEPS model to demonstrate its usefulness in Performance Based Contracting (PBC)
5. Develop a risk model to inform the PBC parties of contractual and financial limits based upon BEPS model predictions
 
CEE FY18-13_Research Framework

Original Research Proposal

CEE FY18-13 Seed Proposal

 


Research Methods

The research builds on the recently completed CIFE report on how design stage BEPS models can be improved in accuracy and the effort to produce them reduced.  One of the major hurdles in this process of building these models is the accuracy of the building geometry.  The Building Parser (BP) project has recently introduced a method to produce accurate internal building geometry  from a laser or Matterport scan.  This project has been designed to test how accurate that geometric data is in terms of energy simulation requirements and how it can be automatically imported to a BEPS system such as EnergyPlus (see Figure 2).

Four buildings will be identified for the research work: 
• Building 1: existing conventional building with an as-built BIM
• Building 2: existing conventional building without an as-built BIM
• Building 3: in the commissioning stage built on the basis of a BIM
• Building 4: in the commissioning stage built without a BIM

The primary purpose of the research is to explore the relative geometric accuracy of the BEM generated by the Building Parser in terms of producing an input for BEPS, versus that of an existing BIM or as-built drawings, depending each time on the building being examined.

 

CEE FY18-13_laserscanToBEM
Figure 2. Workflow of converting a raw point cloud to a suitable BEPS input. The workflow represents the developed pipeline; items in red outline will be validated during this project. All steps are proposed to be automated, apart from the incorporation of thermal and HVAC information.

 


The research method has been divided into five separate sections:

1. Building the hypothesis – the Building Parser method is at least as accurate as the conventional manual methods of importing building geometry to EnergyPlus but significantly more efficient. The steps described below are summarized in Figure 3.
a. Based on conventional methods, build a fully calibrated BEPS model for building #1, using the available geometric data.
b. Repeat the process using geometric data from the Parser output module
c. Check and compare the accuracy of both geometry and BEPS outputs versus actual building and the effort required to complete the BEPS model
d. Refine the method of geometric data export from the Building Parser, if required
e. Repeat this procedure for Building #2 with as-built drawings
f. Build a Best Design BEPS model for building #3, using the available geometric data
g. Repeat the process using geometric data from the Parser output module
h. Check and compare the accuracy of both geometry and BEPS outputs versus actual building and the effort required to complete the BEPS model
i. Refine the method of geometric data export from the Building Parser, if required
j. Repeat this procedure for Building #4 with as-built drawings

 

 

CEE FY18-13_Process
Figure 3. Iterative process to develop and evaluate the proposed approach. These steps will be repeated per building.

 


2. Reporting and refinement
a. Refine the geometric data extract module based on the results found, if necessary
b. Determine the accuracy requirements/capabilities for the Building Parser and BEPS models. To this end, the term “accuracy” in cases of energy simulation and point-cloud generated needs to be formally defined.
c. Publish report and PR Papers

This project provides a significant test of the ability of the Building Parser System to deliver detailed and accurate geometry to a BEPS model and the effect of this approach to obtaining as-built building information on the BEPS model. It will also contribute to assess the accuracy of the developed scan-to-BEPS pipeline that requires minimum human intervention (limited only to the addition of thermal properties).

 

3. Non-geometrical parameters’ BEPS model
The previous study, which utilized data from two buildings in different geographical locations and of very different construction and geometry, resulted in a modelling methodology which provides quantitative guidelines for design stage BEPS models.  The study taught us about the various parametric groups necessary to focus on leading to an accurate yet simple model.  It is now proposed to onwards develop that methodology to include an accurate model using the simplest possible or minimal dataset from the building in question. 
This can be achieved through the following:
a. Collection of utility data and preferably interval data for all energy sources (this can be recorded post project commencement, if necessary)
b. Geometry data collection by laser scan and/or BIM database
c. Examination of material thermal properties from construction specifications
d. Detailed analyses of HVAC systems as implemented versus Ideal loads
e. Best Design BEPS model development
f. Detailed analysis of the probability of using the Ideal Loads model scenario as an accurate predictor of optimized energy usage in both buildings
g. Examine the outcomes for model accuracy and possible improvements to shorten time and effort required

 

4. Assessment of the risks inherent in Performance Based Contracting
More efficient building energy performance contracting requires appropriate allocation of risk.  Appropriate allocation can be achieved by allocating risk to the party involved in the building life-cycle.  The possibility to quantify the influence of BEPS input parameters’ variability on building energy performance will support stakeholders in developing mechanisms to sort out performance responsibilities among designers, owners and building operators.
It opens up possibilities of defining contractual solutions to communicate about anticipated use patterns between owner, building operators, and design team in the design phase. This enable and support the introduction of contractual solutions to allocate risk factors to the stakeholders that can most effectively mitigate the risk. Performance Based Contracting (PBC) requires appropriate allocation of risk within designers, owners and building operators to pursue and ensure building performance delivery. Firstly, the contracting parties agree upon the definition of these risk factors and identify the variables related to each category based on their respective risk mitigation capabilities.  Next, the performance contractor develops a BEPS model which allows the quantification of the impact of these risk factors on the building energy performance through a sensitivity analysis and risk assessment. Finally, the parties agree on a financing structure which determines how incentives / penalties are distributed in the cases where these risks are incurred during building operations.

Risk assessment consists of variating model inputs to see the effects on model outputs in order to determine the relation between independent and dependent variables. The quality of analysis’ result mainly depends on the quality of the models and its input data. BEPS inputs are affected by uncertainties that may have significant effects on outputs and are important to be considered. Having accurate BEPS input improve accuracy of predictions.

 

 

 

Funding Year: 
2019
Stakeholder Categories: 
Owners