There is emphasis across the collections systems industry worldwide for sewer condition prioritization methods and manhole condition ranking. Although currently the UK may be ahead of the US in this field – in the US the focus has historically been on water quality and flow aspects - asset management is now gaining strength in the US.
The concept set out in this paper was developed in the US. Although there are some challenges of nomenclature and terminology in its use outside the US – for example the definition of criticality is slightly different to that of the UK – the concept itself has wide applicability.
The concept, developed by CH2MHill, is available as SCREAM™, a sewer and manhole condition assessment tool. It has been built into Wallingford Software’s InfoNet asset management solution as one of the standard choices for sewer and manhole condition grading.
The pressures for better asset management
One of the main drivers towards better asset management in the US is the framework of Discharge Permits. Every US utility is required to properly operate specified Discharge Permits, which involves an accurate knowledge of the condition of their collection systems. The adherence to these requirements is being more closely scrutinised over time, and in the event of a permit violation, regulators typically demand a condition assessment, involving sanitary and sewer evaluation surveys.
A second factor that increases the need for better knowledge of the system is CMOM (Capacity Management Operation and Maintenance), a set of proposed regulations from the US Environmental Protection Agency. Because of their focus on network performance, these regulations increase the need for asset condition information.
The third driver is the constant growth needs of most networks. As greater demands are placed on the existing capacity, and enhancements are considered, knowledge of the condition of the existing assets is an essential part of planning process.
The need for a specific assessment tool
When CH2MHill first looked for a robust assessment tool to meet the need for closer scrutiny of asset condition for use in the US, there was nothing suitable. For example, the codings in use were not keeping pace with current materials, nor with the level of detection technology, such as CCTV and infra-red examination of sewer pipes. A system was needed for use by a utility as opposed to academics.
The existing US defect aggregation process appeared simplistic and flawed – more an art than a science. There was no standard manual across the US. PAPC (Pipeline Assessment Certification Program – a set of sewer condition classifications devised by WRc, formerly the Water Research Company) is intended to standardize codings, but in practice it falls short of standards needed to match emerging technologies and sewer material enhancements.
In part the development of SCREAM™ was prompted by the shortfalls of existing systems. It is important to consider the purpose of condition codings, which is to determine the sewer or manhole’s existing performance status with its intended design performance and the urgency needed to reinstate its performance status. In developing SCREAM™, a basic consideration covered how prioritisation groups one to five were linked to the rehabilitation process.
The SCREAM™ defect scores are based on scientific research. CH2MHill’s Dr. Vani Kathula conducted her doctoral research on defect degradation that resulted in a sampling of common defects being assigned scores based on their relative severity and remaining life expectancy. A second aim was for all the observed defects to be aggregated and used, with no defect data thrown away – some scoring systems discard some data. A third factor related to degradation over time – the US Transportation Agency learned in researching pavement degradation that degradation is best tracked by non-linear functions, so SCREAM™ uses logarithmic functions.
The next aim was to create a streamlined data process from the field to the collation process. At the extreme in the US, a committee of engineers may look at paper forms and then vote on whether each defect should be graded! So SCREAM™ was designed with a process that curtailed that very time consuming effort. It was also decided to integrate the analysis output of SCREAM™ with subsequent process steps, such as cost effectiveness and the rehabilitation steps. In particular for use in the US, there was a focus on I&I (inflow and infiltration) reduction control methods.
SCREAM™ provides a definitive scoring and ranking process and coding scale. All the codes and all the scores are on a scale of one to a hundred. This allows for a better understanding of assigning relative rates. Since the scores are grounded on research values the sewer prioritization is relative to research values and not the size of the data set. In other words, SCREAM™ scores can be prioritized but the scores may be in ranges that the utility operator has determined are not appropriate for rehabilitation, such as scores below 60 in the 1 to 100 scale. The operator never sees the scale nor assigns a number – that is all undertaken by the tool itself.
SCREAM™ for manholes
SCREAM™ deals with manholes as well as sewers. As CH2MHill was developing SCREAM™, it was noted that whilst there were many methods of collecting data on manholes, none seemed right for aggregating the defects in a manhole. The existing methods of assigning defect scores to components of manholes allowed analysis flexibility, but because of the component materials, the manholes are a little different compared to pipes, and this influences the rehab method. Defects are assigned in SCREAM™ by using the level in the manhole (upper and lower), and the component. The vertical axis in a manhole serves to provide an upper and lower level, and those levels are included in the SCREAM™ algorithm to support a rehab analysis. SCREAM™ allows for scoring the upper level, the lower level and then the overall manhole.
SCREAM™ also includes codes for coatings and linings of manholes, and has assigned specific codes and scores to specific materials of both pipes and manholes – bricks, concrete, etc. This addresses the fact that, for instance, a clay component will degrade differently to concrete, so the scoring mechanism is different.
SCREAM™ can be implemented in a number of different ways for manhole assessments. Some companies still use hard copy, which may then be used as input to the CMMS (maintenance management system), GIS, or a spreadsheet database. Alternatively, the user may have the SCREAM™ logic programmed into software, as happens with InfoNet. SCREAM™ can also reside on a data logger – it is intended to be flexible, available in any type of data collection and analysis, and at any point in the collection and analysis process.
A major advantage of using SCREAM™ on InfoNet is the ease with which InfoNet accesses related data. Some of the early users of SCREAM™ commented that it helps to see pictures of manholes and pipes, so running it on the InfoNet platform has enabled access to images, including CCTV.
The fundamental aim of SCREAM™ is to reduce the scope for subjectivity by the operator. Three pieces of information are recorded when a defect is identified, along with the asset ID. The first is the objective category of the asset – for example, a manhole cover. The second element is the type of defect. Survey teams have to use some judgement here – for example, a crack, corrosion, I&I potential, or wear. The third element- the severity- is the most subjective, but the codes are specified through narrative, definitions, pictures and ideally some training. These three elements, plus the Asset ID, are the only inputs for the operator in the field. They do not have to memorise coding systems, nor assign numbers.
The concept behind SCREAM™’s defect scoring formula is based on established scientific principles and is broken down into a primary equation element and a secondary equation element.
MH Score = max(DS)+ ((100-max(DS)*(A*Weight A+B*Weight B))
DS = Defect Score
A considers the total remaining score for all defects
and B considers total remaining number of defect types
The primary equation element relates to the worst defect of an asset that may have multiple defects. The secondary part takes the remaining defects and considers the total remaining score of all the defects, and the number of defect types, in that pipe or manhole. These two elements of the score can then be weighted by factors that add up to 1 – either weighted evenly at 0.5:0.5, or with a higher weight assigned to one of the components. The balance depends on the preference of the user company. Some SCREAM™ clients prefer rehabbing, and this can be reflected by biasing final scores towards pipes or manholes that have high numbers of defects. Others may prefer that the single most serious defect is dominant in ranking assets.
The final ranking table runs on a scale of 1:100, compared with the older US standard of the ASCE – American Society of Civil Engineers – which is a 1:5 scale. The more detailed scale gives more delineation, enabling better analysis of the ranking list.
Comparison between SCREAM™ and ASCE results
Forty random sample manholes were scored under SCREAM™ on the 1:100 scale as an example of the possible difference in outcome between SCREAM™ and ASCE. The SCREAM™ scores were then converted to a 1:5 scale using a straight line: one to twenty being equal to a grade one, twenty one to forty being grade two, and so on. These results were then compared to ASCE assessments of the same manholes. Of the 40 manholes, 18 showed different results for the two methods, with one extreme being a score of 2 in ASCE and 5 – the top score – in SCREAM™. While there is no absolute rationale to say which is the more accurate result, photographic evidence supported the SCREAM™ scores.
Perhaps the most important conclusion came from graphing the comparative scores and looking at the pattern of difference. The indication is that, with a subjective scoring system, observers tend to over-estimate the good and the bad compared with an objective methodology. SCREAM™ placed more in the centre while the other system identified more for ‘urgent attention’. Although this may not seem highly significant, these errors can translate into millions of dollars more rehab work than a better scoring system would indicate.
Finally, the relationship between condition and criticality: much rehab work is based on criticality – the level of consequences of failure – rather than on condition – the probability of failure. However, the correct approach is to use both consequences and probability, and it is essential to start with a good estimate of the condition.
The aim of devising SCREAM™ was to make available a credible defect rating system based on scientific and mathematical principles, especially in the aggregation of defects, where the industry may be a little lax. The aim is to promote consistent comparisons by reducing scope for operator subjectivity. Comparisons with existing defect rating systems suggest that this objective has been achieved.
Finally, the value of a finer grain scoring system of 1 to 100 is the help in planning annual capital investment programs. If only 1 to 5 is used, a big utility may well have more jobs in the highest category jobs than the budget allows, and no system of identifying the highest priority of these. A finer scoring frame is essential.
This paper was presented by Reggis Rowe of CH2MHill at Wallingford Software’s 2006 International User Conference