1
Workshop 17 - Housing Renewal and Maintenance
Aggregated conditition data bridging the gap between
property management and asset management
Ad Straub
a.straub@tudelft.nl
Paper presented at the ENHR conference "Housing in an expanding Europe:
theory, policy, participation and implementation" Ljubljana, Slovenia
ENHR Ljubljana Workshop 17 Housing Renewal and Maintenance
Aggregated condition data bridging the gap between property
management and asset management
Ad Straub
OTB Research Institute for Housing, Urban and Mobility Studies Delft University of Technology
PO Box 5030 2600 GA Delft The Netherlands Phone: +31 15 27 82769 Telefax: +31 15 27 83450 E-mail: a.straub@tudelft.nl
Abstract
Asset management and property management should be based on objective, reliable information about the performance of buildings and building components.
Technical data collected during a condition survey on-site is needed for strategy formulation and for maintenance planning of each estate. Supplementary technical information is needed for the detailed planning and executing of maintenance work. In 2006 a Dutch Standard for Condition Assessment will be introduced. Well-trained building inspectors should be able to provide property managers with objective, reliable information about performance loss and defects of building components. Maintenance managers can exercise control over the maintenance performance levels and maintenance costs.
Several housing associations want to use the condition data for their asset management too. Aggregated condition data will be used for setting condition targets for assets and for benchmarking. For benchmarking purposes the condition data could be linked to the figures of the Aedex/IPD Social Housing Property Index.
A first analysis of aggregated condition data and maintenance costs of housing estates indicates that the relationship of these facts is not univocal. Aggregated condition marks as well as maintenance costs should be seen as independent indicators for the technical state of housing estates. In this way the data should be used in asset management.
Introduction
defining the desired mix of housing (dwelling types and rent level), analysing the performance of the residential portfolio, defining guidelines for management, acquisition and disposition of the estates in the portfolio (Gruis and Nieboer, 2004). The output of property management is, according to Miles et al. (1996) ‘daily’ administrative, technical and commercial/promotional management.
This paper deals with information of technical management processes and especially planned maintenance, being used for asset management purposes.
Asset management and maintenance
Technical management of the housing stock is in our opinion directly linked to the strategic stock management or asset management (Straub, 2002a; 200b). The policy development takes place top down. See figure 1.
Market strategies can be seen as the outcome of the portfolio-management. A housing association constructs a picture of the composition of the desirable dwelling portfolio and sets up market strategies for product lines, for instance invest in growth or divest. The desired quality of housing complexes and dwellings can depend on the product line to which a complex belongs. Complex and maintenance strategies result from top down formulated market strategies and analyses at operational level. On the operational level analyses of product and exploitation characteristics are done. The ongoing policy might be not appropriate for the forecasted future. Besides maintenance strategies design proposals for major improvements are derived from the complex strategies. The planning and the execution of technical maintenance and technical interventions are considered as property management.
Figure 1: Strategic stock management and technical management (Straub, 2002a) Objectives of aggregation
Condition assessments and maintenance planning are key processes in the process of planned maintenance. Technical data collected during a condition survey on-site is needed for the maintenance planning of each housing estate. The use of condition marks of building components makes the technical status of housing estates transferable between building inspectors and maintenance managers. Maintenance managers can exercise control over the maintenance performance levels and maintenance costs. Is also makes the technical status transferable between the maintenance department and those involved in setting up the asset management.
Market strategies for product lines
Execution work
Design proposals major improvements Maintenance strategies
Planning technical interventions and technical maintenance
Complex strategies
Asset management
Technical property management Market strategies for product lines
Execution work
Design proposals major improvements Maintenance strategies
Planning technical interventions and technical maintenance
Complex strategies
Asset management
Several housing associations want to use the condition data for their asset management too. Aggregated condition data will be used for setting condition targets for assets and for benchmarking purposes. It is a tool to implement the desired differentiation of the technical quality of the housing stock. Housing associations acknowledge that the technical quality is just one aspect of quality (e.g. Straub and Vijverberg, 2004). The technical quality is an important aspect if the housing estate strategy is consolidation. This is the case for larger parts of the housing stock of housing associations. Aggregated condition mark targets can be used for planned maintenance.
For benchmarking purposes the condition data could be linked to the figures of the Aedex/IPD Social Housing Property Index (Vlak, 2005). This index documents the return on investment in housing real estate at the end of 2005 for 57 Dutch housing associations who together manage a portfolio of some 800,000 residential units. This index covers a large part of the portfolio of housing associations. In 2004, 508 housing associations owned more than 2.4 million rental dwellings (CFV, 2005). Dutch housing associations wish to gain insight into their performance as regards their social and commercial activities. Among the members of the Aedex there is a lot of discussion about the (planned) maintenance costs in the index. Costs data without quality data are not reliable. The technical performance of the estates could be linked to the maintenance costs data of the Aedex, making the benchmarking more robust.
Research question
This paper focuses on technical data of building components. It highlights the condition assessment process and condition marks as the key outcome. The main research question is: Which data of the planned maintenance process and especially condition data could be used for asset management purposes?
The paper is based on an analysis of the Dutch Standard for Condition Assessment and a method proposed for the aggregation of condition marks for the use of housing associations. The relationship between aggregated condition marks and planned maintenance costs has been analysed in a case study involving 11 housing estates of one housing association.
This paper is comprised of four parts: condition assessment by housing associations, the aggregation of condition marks, results of the performed analyses of the housing estates and conclusions.
Condition assessment by housing associations
All building components have to contend with performance loss through ageing, use, and calamities. Performance loss is measured in terms of defects ascertained. The defects are registered during a survey.
A condition-dependent approach to planned maintenance leads to a decoupling of quality assessment from the determination of maintenance activities. It also provides possibilities for differing performances of building components and maintenance performance levels (Straub, 2002a). A condition survey is a tool in assessing the technical performance of the properties to underpin the long-term maintenance expectations. The information is meant at first for strategic property management. A condition assessment is not meant for preparing the year maintenance budget and planning of the work. Supplementary information is needed in the phase of preparing execution of remedial work.
Dutch Standard for Condition Assessment
The practise of condition assessment by building inspectors yielded variable results due to subjective perceptions of inspectors. A lot of research has been done to create objectivity in the inspection process that should result in unambiguous information for maintenance strategies. As a result of research projects and the use of the method in the Dutch Qualitative Housing Registration (Kwalitatieve WoningRegistratie - KWR), the process of condition assessment using standard lists of defects and a six-point scale has become popular by property managers of housing and real estate, consultants and contractors in the Netherlands. But, the condition assessment methods vary for the hierarchical classification of building components, classified defects and the use of condition parameters: type, intensity and extent. Several condition assessment methods lead to variable resulting condition marks, whilst examining the same defects. The different results of various methods are not a drawback in practice. Important is that within an organisation all building inspectors handle they’re own method the same way. However, it is a handicap in the transfer of people and knowledge between property managers, consultants and maintenance contractors. Standardization will uniform this. Building inspectors can be trained in using one method for all principals. It makes the data also suitable for benchmarking purposes.
In 2002 the Dutch Government Buildings Agency took the initiative to standardise the condition assessment of building components construction, including building services. The objective of this standard is an objective assessment of the technical quality, to provide property managers with objective, reliable information about the technical quality based on assessed defects (performance loss) (NEN, 2006).
The well-known six-point scale is the basis of the standardised method. The condition categories are of a chronological order that describe possibly occurring defects without references to remedial work. Table 1 gives general descriptions of condition marks.
Value General condition description
1 Excellent 2 Good 3 Fair 4 Poor 5 Bad 6 Very bad
Table 1: Six-point scale Dutch Standard for Condition Assessment (NEN, 2006)
An important observation is the scale division. The six-point scale is not linear but ordinal. An ordinal scale division means that the values the variable can have can be classified, but their meaning is not univocal. A building component in condition 3 does not mean 3 times being worse than a component in condition 1. Condition 1 indicates the upper value of the scale. This absolute value cannot be exceeded. Condition 5 indicates the lower value of the scale. This bad condition is not an absolute value. Condition 6 has been added to distinguish a very bad situation, meaning that the component should already have been replaced. A linear scale division presupposes a linear relationship between the conditions and the remaining life span (service life) of the building components. In reality performance loss and life span of discrete building components and sets will run differently.
The condition assessment process follows the pattern in figure 2. The assessing of defects occurs first. Without this information one could not formulate maintenance activities and estimate costs. A visual inspection will usually suffice. Subsequently the inspector passes through the following condition parameters: type of defects, intensity of defects and extent of defects.
Figure 2: Condition assessment process (Straub, 2002a)
For an objective visual assessment building inspectors need a clearly defined and hierarchical classification of building components. This hierarchical classification directly influences the classification of the type of defects. It should be related to possible maintenance activities: maintenance cost elements. For instance on the level of flat roof finishes, e.g. bituminous roll, or roof drainage (gutters and drainpipes) one could assess defects with a certain extent and intensity. Building inspectors are not able to assess defects on a higher hierarchical level, like flat roofs including all finishes and secondary components.
Type of defects
The type or seriousness of the defect indicates to what extent it influences the functioning of building components. The Dutch Standard for Condition Assessment standard classifies the type of defects of distinct building components with minor, serious and critical. Critical defects significantly threaten the function of the building component. Generally material intrinsic defects like corrosion and wood rot, defects that threaten the building structure, e.g. stability and distortion, and ‘functional defects’, are weighted as critical defects. Functional defects are those that are already associated with a breakdown in performance: symptoms. Serious defects are gradually damaging the performance of building components, for example defects in the material surface. Defects to secondary components, for example coatings, are classified as minor defects.
Intensity of defects
The intensity of defects strongly influences the condition of building components. The intensity of defects deals with the degradation process. Ageing defects like material intrinsic defects and defects involving the material surface, e.g. wear and soiling, develop over a certain period and will occur in several intensities. But defects caused by calamities, for instance glass breakage, just occur in one stage. One may register the frequency of defects in
stead of intensity. The frequency of failures is particularly a useful condition parameter for condition assessment in the case of building services components. The Dutch standard uses three intensity classes. See table 2.
Intensity class Name Description
Intensity 1 Low The defect is hardly visible
Intensity 2 Middle The defect is progressing
Intensity 3 High The defect cannot progress any further
Table 2: Classification of intensity of defects (NEN, 2006)
Extent of defects
Besides knowledge about the intensity of defects knowledge about the extent of defects is needed to assess the condition.. Methodological questions rise how many classes be manageable for building inspectors and how many classes be useful to link maintenance activities to the extent of defects in the policy making process. Clearly, to estimate the extent and chose for the appropriate class is difficult, even for more experienced building inspectors. Obviously difficulties doing so also depend on the defect involved. One may differentiate general ageing defects normally covering the whole building component from localised defects. In the case of general ageing defects the intensity of a defect corresponds with the condition. The Dutch Standard for Condition Assessment standard distinguishes five extent classes. See table 3.
Extent class Percentage Description
Extent 1 < 2% The defect occurs incidentally
Extent 2 2% -10% The defects occurs locally
Extent 3 10%- 30% The defects occurs regularly Extent 4 30% - 70% The defects occurs frequently
Extent 5 > 70% The defect occurs generally
Table 3: Classification of extent of defects (NEN, 2006)
Condition marks
The extent and the intensity of a defect combined with the type of the defect lead to a condition mark, probably with a defect score as an intermediary product. Table 4 gives an example for critical defects.
Extent
Intensity < 2% 2% - 10% 10% - 30% 30% - 70% > 70%
1 Low 1 1 2 3 4
2 Middle 1 2 3 4 5
3 High 2 3 4 5 6
Condition targets
Property managers are able to set condition targets of assets, also named maintenance performance levels, by forecasting the condition state of components after executing maintenance activities, dealing with more and less acceptable remaining defects. See figure 3.
Figure 3: Condition assessment process after execution of maintenance work (Straub, 2002a)
Aggregating condition marks
Aggregating condition marks means weighing the technical state of a building component against the other building component of a housing estate.
In the proposed method for housing associations first the weighted average of building components being part of a main component is determined. Main components are for example facades, roofs, windows and doors, balconies and walkways and heating. The weights are diverted from the hierarchical levels of building components used in the Dutch Qualitative Housing Registration 2000 (KWR 2000). This means that construction parts count for 4, finishing and equipment parts count for 2 and painting counts for 1. For example in case of facades the condition of the brickwork counts for 4, the condition of exterior staircases (equipment) and plasterwork (finishing) count for 2 and the condition of the painting counts for 1.
Secondly the weighted average of main components of the housing estate (housing block) is determined. The weights are diverted from the financial share in the replacement costs of the building, based on data of the KWR 2000.
Case study
In a case study technical data and costs data of 11 housing estates of one housing association has been analysed and compared. Maintenance activities and maintenance costs in the long-term maintenance planning have been based on a condition assessment using condition marks for building components. For the aggregated condition mark of a housing estate the aggregation method as described above has been used.
A first analysis indicates that the relationship between the aggregated condition marks and the maintenance costs recorded in the long-term maintenance planning is not univocal. Also the relationship between the maintenance costs and performance improvement - aggregated condition marks before and after maintenance - is not clear. One reason for this could be the fact that the aggregation of condition marks of building components comprise all building components, including components that do not need maintenance work within 10, 20 or even 30 years. For example generally occurring ageing defects with a low intensity will lead to a condition mark 3.
Unfortunately the case study is unfinished. Maintenance costs data from the long-term maintenance planning is somewhat unreliable for some buildings components, especially the interior components and the HVAC. Besides, the maintenance planning covers just ten years. Another drawback is the unawareness of the status of the maintenance planning. Did asset policy considerations already effect the maintenance planning? Further research will be done to make the planning more reliable. The same analyses will be made in a second case study. Henceforth research to the dispersion of condition marks per year of construction and type of dwelling is needed. The year of construction is particularly relevant in that it reflects the building regulations in force at the time as well as the financial standards for subsidised public housing, technological developments, the methods of construction employed and the general economic climate. The year of construction also reflects significant differences in present technical condition.
Conclusions
The asset management and maintenance policy-making should be based on objective, reliable information about the performance of building components. Data are required on the technical condition of the building components (e.g. condition marks), the housing quality (e.g. standard of equipment and finishing of the kitchen), the environmental quality (e.g. use of materials, energy-use and kind of heating system), adaptability for changes in housing and environmental quality and expenditures for maintenance and improvement.
Condition-based building maintenance using the Dutch Standard for Condition Assessment gives a useful tool for inspection and planning maintenance work. Building inspectors can provide property managers with objective data about performance loss and defects of building components. It is to be expected that as a result of the Dutch Standard for Condition Assessment the process of condition assessment using standard lists of defects and the six-point scale will become more popular among housing associations.
For benchmarking purposes aggregated condition marks could be used. Aggregated condition marks as well as maintenance costs should be seen as independent indicators for the technical state of housing estates. To underpin the asset management by technical data, the following indicators could be used:
• assessed condition marks building components before maintenance
• condition marks building components after maintenance (calculated or estimated) • aggregated condition marks main components
• aggregated condition marks housing estate (C aggregate) • target condition marks main components (C target) • target condition mark housing estate (housing blocks) • maintenance costs main components
• total maintenance costs housing estate coming 30 years (long-term maintenance planning)
• total maintenance and improvement expenditures last ten years
• relative maintenance costs (maintenance costs compared to market value)
Strategies Housing portfolio
Consolidate Invest Divert Total Standard Number of dwellings: …
C aggregated: … C target: …
Luxurious
Live and care Product line
Total
Table 5: Example asset management using condition data
Acknowledgements
Consultant PRC in co-operation with housing association Ymere conducted the data analysis of aggregated condition marks in the case study. The author thanks Hajé van Egmond from PRC, Pablo van der Laan from Ymere and Arie Warnaar from consultant ICP+Consult for the analysis and the discussion over this.
References
Centraal Fonds Volkshuisvesting (CFV) (2005) Verslag financieel toezicht woningcorporaties
2005, CFV, Naarden, The Netherlands.
Gruis, V. and N. Nieboer (2004) Strategic housing management. An asset management model for social landlords. Property Management 22(3), 201-213.
Miles, M.E., Haney, R.L. and Berens, G. (1996) Real estate Development, Principles and
Process, ULI, Washington, DC.
NEN (2006) Ontwerp NEN 2767 Conditiemeting van bouw- en installatiedelen, NEN, Delft, The Netherlands.
Priemus, H. (2003) Dutch Housing Associations: Current Developments and Debates.
Housing Studies, 18(3), 327-351.
Straub, A. (2002a) Strategic Technical management of housing stock: lessons from Dutch housing associations. Building Research & Information 30(5), 372-381.
Straub, A. (2002b) A framework for strategic technical management. Proceedings ENHR
conference Vienna 2002 Housing Cultures: Convergence and Diversity, ENHR/Europaforum,
Wien, Austria.
Straub, A. (2003) Using a condition-dependent approach to maintenance to control costs and performances. Journal of Facilities Management, 1(4), 380-395.
Straub, A. (2004) Housing management and maintenance practise of Dutch housing associations. Proceedings ENHR conference Cambridge 2004 Housing: Growth and
Regeneration, University of Cambridge, UK.
Vijverberg, G. (2005) Technisch beheer en strategisch voorraadbeleid in de
corporatiepraktijk; Uitkomsten van een sectorbrede enquête, Habiforum, Gouda, The
Netherlands.
Vlak, A.L.M. (2005) Shedding light on the performance of housing corporations. Holland
