Mind the gap!
The right design for heat transfer units

Organic heat transfer fluids have become virtually indispensable in many processes. As a rule, their use is unproblematical due to low pressure even at high temperatures. However, to optimize the service life of heat transfer units, several points need to be observed. Support with the design of such a unit can be obtained both from specialized companies and from experienced heat transfer fluid manufacturers. This article describes some of the stumbling blocks that are overlooked even by professional system planners.

Heating and cooling play a key role in the control of chemical reactions. Heat transfer units thus operate, to a certain extent, “at the very heart” of industrial production plants, and their reliability is absolutely essential for safe working. Nevertheless, even the operation of heat transfer units is now being viewed against the background of the growing cost pressure to which the chemical and petrochemical industry has become increasingly subjected in the last few years.
It is common knowledge that organic heat transfer media – which have a whole number of technical advantages over water – gradually decompose at high service temperatures (Fig. 1). Nevertheless, the process engineer nowadays would expect a heat transfer fluid to perform its job reliably in such units for at least a year without being replaced. In fact, this sort of service life can generally be guaranteed without difficulty, but only if the heat transfer fluid has been correctly selected and the unit has been designed according to the state of the art. After all, the course for long-term, trouble-free operation of a heat transfer unit is determined as early as the design stage. Errors made in the design wreak revenge later in the form of above-average downtimes.

Studying the product data is not enough
Even choosing the most suitable heat transfer fluid is more complex than simply comparing values in a table with the intended operating temperature of the pro-
jected plant. The heat transfer market has a whole number of alternatives designed for service temperatures up to 400 °C. In particular, non-substituted aromatic hydrocarbons like the members of the Diphyl family from Lanxess Deutschland, have particularly high stability. (Lanxess is the name of the new company that comprises the vast majority of Bayer’s chemical activities plus about a third of its former polymers business.) Here, the inherent stability of the chemical bonds in aromatic molecules ensures very high decomposition temperatures or very slow decomposition reactions (see box on next page). Apart from this, there are a whole number of other points to be taken into account if the heat transfer unit is to operate economically and safely over a long period of time. Before deciding on a heat transfer medium, all the advantages and disadvantages of a product should be very carefully weighed up. For example, there are some fluids that have very high heat stability but at the same time have disadvantages in terms of their viscosity or reprocessing properties. Some also have a tendency to attack seals and gaskets, and, in the event of a leakage, do not give off any odor.
Low pressure guarantees safe operation
The design of the heat transfer unit itself also entails a number of stumbling blocks. As a rule, these units can be operated under relatively low pressure, which means that the energy stored in them is relatively low and constitutes only a minor potential hazard. Nevertheless, there are a number of factors that have to be taken into account. Being organic compounds, heat transfer fluids support combustion, even though the auto-ignition temperatures of most organic heat transfer fluids are very high. This aspect must, of course, be taken into account when covering the pipes with insulating material. Most of the materials used for insulation have a large surface area, which means that slow exothermal reactions between the atmospheric oxygen and escaping organic fluids can occur in the heat insulation, possibly over a large surface area, and can very gradually heat the insulating material above the autoignition temperature, possibly even resulting in a jet of flame.
On the other hand, such problems can be solved by taking suitable measures at the design stage, often merely by attending to minor details. It must, for example, be possible to recognize any leakage in the pipe system from outside at any time. Heat-resistant graphite seals with a steel insert can significantly increase the safety of the piping system.
It is also particularly important to bear in mind that although the pressure testing of a unit – performing a hydrostatic test, for instance – may say a lot about its strength, it says nothing about its leakproofness! Ideally, this should be monitored using special methods like the helium test, and preferably for the whole pipeline system, which must naturally be designed to take into account the expected changes in length at the intended service temperature. A steel pipe, for example, expands by about 4 mm per meter at a temperature difference of 300 K (Fig. 2). Corresponding additional loads for the pipe connections and claws of the apparatus and for the pipelines themselves must also be considered in the design. In practice, it has been found that failure to perform stress tests on the pipelines can lead to considerable problems in subsequent operation.
On the other hand, the dreaded gradual thermal decomposition of the heat transfer fluid is much less harmful that is often assumed. It is dependent exclusively on the temperature of the medium and can therefore, in principle, be controlled – as long as it is not catalyzed (accelerated) by, for example, contamination entrained while setting up the system. In such cases, a thorough quality control should be carried out (Fig. 3)! As a rule of thumb, it can be said that, under normal circumstances, an increase in temperature of 10 °C doubles the decomposition rate of the heat transfer fluid. Logically then, the service temperature of a heat transfer unit should always be kept as low as possible. Reducing it by 10 °C is enough to have an effect. On the other hand, this dependence on temperature also means that sections of the plant in which the heat transfer medium heats up unnoticed to higher temperatures than planned can lead to a considerable accumulation of decomposition products. There are, of course, various procedures for dealing with this. Low-boiling decomposition products are relatively unproblematical. Although they can lead to pressure fluctuations in the system, they can be expelled by desorption or stripping. Contamination from high-boiling products is more difficult. Although it can be eliminated with by-pass filters, it can in extreme cases form deposits and block the system, hindering the heat transfer process. To restrict the formation of these decomposition products and avoid premature replacement of the heat transfer fluid, it is not enough simply to keep a general watch on the starting temperature and exit temperature. The decisive factor here is the film temperature on the heating rod or at the tubes inside a boiler! There have been several occasions in practice, in systems of parallel heating coils, in which temperature differences of many dozens of Kelvin combined with appreciable local overheating have occurred because inadequate attention was paid to the flow conditions, even though the measured starting temperature was always “okay” (Fig. 4). After all, the transfer of heat via the surface of a heating rod or tube takes place in a liquid boundary layer that is no more than about four tenths of a millimeter thick. Should local overheating occur here – and this is generally impossible to tell from the starting temperature – the result can be a dramatic increase in the decomposition of the medium, and even carbonization. As we know, carbon is a good insulator, which means that such deposits can also cause considerable damage due to overheating of the heating elements.
Even sophisticated flow calculations, which reach their limitations especially with complex pipe or heating rod geometries, are not always able to tackle local overheating of this kind. This is why experience and expertise are so important! As a general rule one can say that corners in which low-flow zones can form should be avoided and sufficient safety margins should be provided.
Important: Keep an eye on the viscosity
One other point that is unfortunately also often neglected is the viscosity of the heat transfer fluid. If a conventional heat transfer pump is used, care should be taken that the viscosity of the liquid medium is always below approx. 5 to 10 mm2/s. It is important to bear in mind here that even though this condition may be satisfied when the plant is in operation, it is not necessarily so when the plant is started up cold. It may well be necessary to make use of additional heating systems for the start-up phase. There are also several points to bear in mind when selecting a pump. Where the temperature is likely to reach around 300 °C and above, use should always be made of canned motor pumps or magnet-coupled pumps. Although they are significantly more expensive than conventional pumps, they have much lower leakage rates at elevated temperatures than pumps with axial face seals, and are thus the more economical alternative because they reduce plant downtimes.
But a high viscosity is a problem not only for the pumps, it also significantly impairs the heat transfer process and thus increases the risk of overheating on the heating element. Frequent unnoticed overheating during the start-up phase can considerably accelerate the “natural” decomposition reactions and lead to the formation of deposits that harm the system and reduce the service life of the filling. It must also be taken into account in this connection that the viscosity of an organic heat transfer fluid can increase significantly through the formation of high-boiling decomposition products at service temperatures above 280 °C. Where air is present, 60 to 100 °C may even be sufficient. This can lead to the formation of organic acids, which, in the long term, can also result in corrosion of the pipe system. By regularly monitoring the neutralization index of the heat transfer fluid, such phenomena can be identified.
For this reason, both DIN 4754 (German Standard for heat transfer units) and the new European Operational Safety Regulations specify annual checking of the heat transfer fluid. Reputable heat transfer media manufacturers generally offer a comprehensive analysis service through which – provided they have sufficient experience – specific problems with heat transfer units (e.g. due to infiltration of product into the heating system) can be quickly solved.
Another aspect of quality assurance is to involve the company supplying the plant components. It cannot necessarily be assumed that all such companies are familiar with the specific problems involved in operating a heat transfer unit, nor can it be taken for granted that they have the necessary know-how about the material properties or, for that matter, adequate testing facilities. However, an experienced engineer is able to define clear interfaces between the various service providers, thereby eliminating many potential problems before they happen. One example is the specialized engineering company HTT energy systems of Herford, Germany, that have such know-how. So, too, do various highly competent manufacturers of heat transfer fluids, like Lanxess Germany.