84 km of the replaced double pipes are distribution and connection pipes, which means that approx. 13% of the total examined network has been replaced. Table 2 indicates the extent of replacement in the individual areas.
In many district heating (DH) networks age has begun to set in and there is now a pressing need for a thorough renovation.
Of Denmark's DH facilities, about 275 were built before 1970 and time is catching up on many installations which originally were established with a life expectancy of some 35 years.
This chapter deals with (1) renovation of older systems, illustrated by a description of activities carried out in Odense Municipality. and (2) treatment of circulation water which is very important for attaining high efficiency in the system and avoiding damage to the system.
The age problem of an installation is generally evident when the number of pipeline breakdowns per kilometre begins to escalate. The resulting energy loss is most apparent in the summer when consumers exclusively use the system to supply hot water. When the district heating system has reached a state that necessitates renovation it is not unusual to observe an energy loss as high as 40% resources which simply disappear into the ground.
The majority of the old networks suffer because of their location in concrete pipe ducts. In many cases the pipelines are supported on metal bearers and where these are in physical contact with the mains piping. the insulation becomes damaged due to pipe movement. It is at these damage points that serious heat losses occur.
But the worst of the problems connected with the concrete ducts is dampness. When the insulation as a rule, mineral wool or cellular concrete becomes damp its insulating ability drops sharply. Additionally, the combination of moisture and heat further increases the risk of corrosion. If rust sets in at some point, it quickly spreads under the insulation material.
In networks in concrete ducts, localising leakages is extremely difficult and expensive. Apart from the extensive excavation required to get to the pipeline, the concrete casing has to be broken into and reestablished afterwards, making concrete work the most expensive part of this kind of DH installation repairs are therefore carried out only in the event of major breakdowns. For this reason network maintenance is not always sufficient and the network gradually deteriorates more and more.
The replacements for pipes in concrete ducts are preinsulated DH pipeline systems. By means of computer technology, monitoring systems of this kind can be centralised, and should breakdowns occur, damage points are easily located. The quality of preinsulated piping, especially the insulation,is much higher than in networks, with concrete ducts.
In many locations in Denmark it has been decided that renovations of DH networks are to be carried out continuously. Thus, in stead of a total hereandnow replacement of concrete ducts, efforts are concentrated on those parts of the network most exposed to damage. e.g. where pipe runs lie in damp or clayrich soils. In some locations the concrete ducts have not been removed at all, but instead a completely new network of preinsulated piping is laid on top of, or at the side of, the old network.
In many localities the renovation work has brought about extra savings as new dimensioning standards for the systems were established. The old pipelines were often overdimensioned, resulting in unnecessary heat loss.
Ever since the modest start in the late 1920s the central parts of Odense have been provided with district heating. Initially, the heat was supplied from the power station located in the centre of the city.
In 1953 a new combined heat and power station "Fynsvaerket" was inaugurated, which permitted the DHsupplied areas to be enlarged.
The experiences gained from Odense municipality's DH scheme were disseminated to the surrounding districts, where a number of private and municipal DH boards were formed. The heat for these outlying systems was supplied by oilfired DH stations.
As a result of the municipal amalgamation of 1970 an these DH boards were brought together under a single administration and the Individual DH pipe networks linked up with the central Odense transmission system. Thus. "Fynsvaerket" took over an increasing share of the DH In Odense municipality.
The DH scheme was extended until 1988. at which time 95% of the Odense municipal district had been covered. The DH transmission network consisted of 1,500 km double pipes with more than 50,000 connections.
However, the quality of the total transmission system varied from one section to the other. The following materials had been used :
-Concrete ducts with aerated concrete insulation, cast on site
-Concrete ducts with asphalt insulation, cast on site
-Prefab concrete ducts with aerated concrete insulation
-lst generation, preinsulated pipes
-2nd generation. preinsulated pipes.
Concrete duct pipes and lst generation preinsulated pipes totalled 650 km distribution pipes and 27 km transmission pipes.
The number of leaks from these pipes was substantial and they occurred at a growing frequency with no clearly recognisable pat tern.
The local authorities therefore requested a consulting engineering company to perform a technical/economic rehabilitation study. The objective of the study was to identify the best possible way to minimise the frequency of damage, subject to keeping within the resource constraint of the DH Board's budget.
The DH area mentioned above was divided into 18 subareas of a manageable size and the investigations of individual streets could be related to these subareas in later detailed studies. The boundaries between subareas were chosen so that they followed main traffic arteries or other features of the Odense townscape, such as the small Odense River (see figure 1). For each of the areas all pipes were measured, connections were counted, and the type of pipes recorded. All these data were stored in a computer so as to permit sorting according to dimensions in a given area.
Based on the above and on the technical and economic information available, computations were made for:
-Thermal loss in existing pipe network per year
-Thermal loss in new pipe network per year
-Difference in thermal loss per year
-Reduced cost of thermal loss per year
-Cost of excavating pipe network for rehabilitation
-Cost of reestablishing pipe network
-Total costs of rehabilitallon.
To evaluate the profitabillly of the rehabilitation for the individual areas, all cash inflows and outflows associated with the investment were computed for a lifetime of 20 years, a period believed to be short compared to the actual lifetime of these systems
Cash outflow for the 20-year period was:
-Rehabilitation cost.
Cash inflows (reduced outflows) over the period were:
-Savings on reduced thermal loss
-Savings on reduced water leakage
-Savings on reduced pumping energy due to reduced flow as a consequence of reduced thermal
loss
-Saved costs for repair works.
It was considered important in the investigations to be able to form a picture of how the frequency of leaks would develop over time.
In the period 19781987 approximately 3000 leaks were recorded for the DH pipe network and the frequency was clearly increasing over the period.
In the analysis the various kinds of damage to the pipe system were sorted on the basis of locality and year. In this way it was fairly easy to register the extent of the damage in a specific geographical area.
The extent of the damage in each of the 18 areas can be expressed graphically as a function of time using linear regression.
The expression for damage in the period 19771987 can be expressed:
y = bo + blX
where bo is the theoretical number of leaks in 1977 and b1 the slope, an indicator of the increased number of leaks.
The theoretical number of leaks in 1977 is a computational value that is used when projecting the number of leaks in the period considered.
Profitability was evaluated on the basis of recorded costs and revenue. As mentioned above, the costs are the Initial costs of the rehabilitation, whereas the revenue takes the form of saved expenses for. e.g. repair work. These cash flows were discounted to the present. i.e. to the beginning of the period considered, using a discount rate equal to the real rate of Interest In the financial market.
The criterion chosen for the project was the internal rate of return, which is defined by the following expression:
where "n" Is the length of the period considered and "I" the internal rate of return. The expression says that the sum of all revenues minus the sum of all expenses dis counted to the beginning of the period. using the internal rate of return, equals zero
Table 1 shows the result of the analysis. The areas most in need of rehabilitation are listed at the top of the table. For some of the areas the net present value of the investment is negative. i.e. the investment is not profitable as the internal rate of return is lower than the discount rate.
In connection with the analysis, sensitivity analyses were made for an increase in initial investment, increased energy prices, and other more technical parameters. An in crease in the rehabilitation costs of. e.g. 10% would result in a fall of the internal rate of return in the different areas of 1.2 3.l%. All other changes of parameters resulted in higher profitability.
In connection with the rehabilitation works it proved necessary to develop computerbased programs for pipe network calculations, as the present load did not at all correspond to the consumption patterns used as design bases when the network was originally constructed.
The consumers are more conscious about how they use the resources, houses and buildings have been modernised, insulated, and provided with double glazing, and the supply possibilities have changed. All this means that substantially larger areas can now be supplied without requiring larger pipe dimensions in the distribution network.
The Odense Municipal DH Board prepares rehabilitation projects for 4050 km of pipe network every year. The works are carried out by local contractors who are chosen on the basis of tenders for longterm contracts (typically 4 years). Over the years these contractors have worked closely together with Odense municipality and they have a thorough knowledge of local conditions, which makes their prices absolutely competitive.
Where possible, it is attempted to place new networks in new alignments in order to make the rehabilitation cheaper. Another advantage of this procedure is that the networks can be constructed in stages and planned so that the consumers along a stretch to be rehabilitated will be cut off from the system for only about 812 hours. depending on the scope.
When rehabilitating, both main and connection pipes all the way up to the consumers' main cocks are replaced so that the users will have a totally upgraded supply.
After 5 years of rehabilitation with an annual budget of DKK 3550 mill. Odense Municipal DH Board wanted an appraisal and a revaluation of the 1988 report in order to determine whether there might be reason to change the priorities given to the rehabilitation works in the 1988 report.
Since 1988 Odense Municipal DH Board has replaced a total of approx. 103 km double pipes in the DH pipe network, of which approx. 51 km have been replaced as systematic rehabilitation and the remainder in connection with repair works.
The reason that a relatively large number of pipes approx. half have been replaced in connection with damage repairworks is that it has proved to be economically advantageous to replace a larger stretch of pipe than immediately necessary in connection with damage repair works.
84 km of the replaced double pipes are distribution and connection pipes, which means that approx. 13% of the total examined network has been replaced. Table 2 indicates the extent of replacement in the individual areas.
The effect of the rehabilitation works on the pipe network are best seen by examining the development in damage frequencies.
Figure 2 shows that the frequency is clearly increasing in the period 19781987 and slightly declining during 19881991.
Although the frequency only falls moderately after a systematic rehabilitation of the pipe network has begun, it can be concluded that the rehabilitation has had a positive influence on the damage frequency even though no more than about 13% of the total of 646 km distribution and connection pipes have been replaced.
Based on a closer analysis of damage frequencies in the individual areas, it can further be concluded that approx. 2025% or more of the pipe network must be replaced before the frequency is reduced markedly. 
The examination of pipes showed that 563 km double pipes needed replacement at a total rehabilitation cost of DKK 821 mill. However, it would be economically advantageous at the present time to rehabilitate only 442 km at a total cost of DKK 627 mill, (see tables 3 and 4).
It can also be concluded that there is no reason to make significant changes in the priorities allocated to the rehabilitation works in the 1988 report. Only slight shifts in ranking have been made.
It will be necessary to continuously monitor the areas which are not ranked high in the list of priorities, as a sudden increase in the occurrence of pipe damage in a certain area can completely change the priority that should be given to this area in the rehabilitation programme.
As it appears from the above, it is essential to perform a thorough analysis of the entire DH system before rehabilitation works of the pipe network begin. There are many parameters that may be important to achieve a good result, both for the economy of the actual rehabilitation and for the subsequent period of operation.
For example, it might be an advantage to start by improving the insulation of buildings as well as the possibilities of controlling the internal heating systems. Thus, by reducing the need of heating energy for the buildings, the supply capacity of the pipe network can be reduced accordingly, which results in smaller dimensions, and lower rehabilitation costs, reduced requirements to the DH stations, etc.
So, careful and thorough preparatory studies are preconditions for insuring success for the entire rehabilitation scheme.
Throughout the history of DH many plants have been subject to energywasting formation of scale or of destructive corrosion of the plant. This is caused either by lack of or inadequate water treatment.
With the object of minimising these risks, guides have been compiled describing appropriate water treatment. DH plants should thus be filled and tempered with softened or totally desalinated water, and pH values should be adjusted to between 9.5 and 9.8. But there are also other important factors concerning water treatment, and these necessitate regular analysis and readjustment of the circulation water of the DH plant.
There is, for example, a constant risk of hard infiltrations entering the system via leakages, as from hot water tanks. At the start this will lead to a buildup of sludge or deposited scale, the latter occurring in the pipe network or on heat transfer surfaces, in the boiler pipework or in the heat ex changers. This will greatly influence operating costs, as only 1 mm of scale is enough to increase fuel costs by 10%. The presence of oil or grease in the water in solution with lime sludge will cause an insulating scale to be formed.
Corrosion in DH systems can be minimised primarily by tempering the system with water that has passed through a deaerator, thus reducing the concentration of dissolved oxygen to below 0.02 mg/L. Additionally, it is essential that chemical dosing be strictly controlled. As a rule, as noted above, the pH value should be adjusted to between 9.5 and 9.8. This value will ensure that the selfprotective magnetite layer of the iron will remain intact, that the corrosion of brass will be at a minimum, and the builtup copper oxide coating will also remain intact. However, consideration should also be given to the infiltration of oxygen which occurs, and an oxygenabsorbing chemical should be added. Lastly, it is appropriate to add a dispersing agent to the water to ensure that possible sludge formation is easily freed and may thus be filtered.
In subsystems such as heat exchangers containing stainless steel the chloride ion concentration must not exceed 100 mg/L or galvanic corrosion will occur. Attention should be given to the possible buildup of microbiological growths these will initially produce unpleasant odours through the formation of hydrogen sulphide. The presence of iron and sulphatereducing bacteria is harmful to the system and when fully developed can lead to serious corrosion damage. These can be prevented only by timely intervention with a disinfecting treatment.
Water treatment will always cause lime or ferric sludge to form, and either will lead to wear and tear and operating problems if they are permitted to continue circulating. It is therefore advisable that approx. 5% of the circulating return water be continuously filtered in a combined magnetic and bag filter.
Application of the above measures will en sure that DH water is clear and free of odours and sludge, these are the preconditions for a long plant life.