GIS based decision support for gas pipe maintenance

3 Architecture of a SDSS for pipe maintenance
A typical architecture of a SDSS consists of five key modules (Armstrong et. al., 1990):

  1. data base system
  2. model base system
  3. display generator
  4. report generator
  5. user interface
A typical GIS as used today is centred around a spatial database system (enabling spatial queries and spatial access methods) and a user interface. Functionality extensions for typical business lines of a utility company like natural gas or water can be set up on top of the database.

For the SDSS used in the context of gas pipe maintenance a three layer architecture can be used as indicated in figure 2.


Figure 2: Architecture of Spatial Decision Support System


The SDSS blends with the existing GIS and utilizes all resources and GUIs the GIS offers. The model base for typical procedures that supports the decisions in pipe maintenance has to be developed.

4 Model base construction
In the field of gas pipe maintenance there are four fields of decision making. All of them carry spatial reference and are closely interrelated:
  • Estimation of current state of decay of each pipe segment, depending on material or age, but also on traffic load, tree cover or building proximity
  • Choice of maintenance technology which is influenced by environmental restrictions like traffic load or number of consumer stubs
  • Cost incurred, depending on various cost drivers
  • Site pooling, considering the cost saving perspective and the applicability of mainte¬nan¬ce methods
The supposed model base has four core components as outlined in figure 3:

  • a system to measure and estimate the state of decay or each segment
  • a component to identify technically feasible refurbishing techniques for each pipe and each construction side
  • a cost module
  • and a cluster component to aggregate construction sides.
Immediate action leaks are handled as special cases. If the leaks are found within an already iden¬tified cluster, the whole cluster is given immediate treatment, else the large leak is hand¬led in isolation.


Figure 3: Concept of the model base


4.1 Scoring
Urgency of preventive maintenance is measured trough a scoring model. Typical relevant influencing factors are technical data (pipe diameter, building distance), economic factors like outdated pipe material and external criteria as traffic load or unapproved tree cover (DVGW, 1999). For each criteria categories are formed and a score is assigned to each of them. The pipe indicated in figure 4 earns 5 points for being small diameter, 2 for the soil type and so on. Summing all categories, there is a total of 16 urgency points. By using this score different line segments can be compared. This allows for ranking the whole distribution network on a common perspective. Most of the influence factors (such as the ones indicated on figure 4) can be taken from a spatial database immediately. The ultimate model within a company has to be geared to what data is available.


Figure 4: Scoring for an exemplary pipe segment


4.2 Selection of technology
Second part of the model base is the choice for a good maintenance technique. These techniques can be classified as repair, reconditioning or replacement measures. Repair is concentrated on leaky junctions and bushings. Reconditioning techniques try to refit existing pipes with a new inner lining for instance through hose relining. Replacement can be done with traditional open pit technology but also with trenchless methods. Only two small pits are needed to destroy the pipe by a burster to create place for a new pipe.

Each of the technologies carries preconditions for applicability, for instance soil conditions, tube material or the number of stub connections, which can be found within the spatial database. Now, given the characteristics of the line segment, infeasible techniques are filtered out in step 1 (figure 5). The remaining alternatives are now subjected to cost calculation. In preparation to cost calculation the priority metric which is calculated as urgency score points resolved divided by the cost incurred was especially useful. The number indicates where the invested money is spent in a most efficient manner.


Figure 5: Choice of maintenance technique


Page 2 of 3
| Previous | Next |