CC BY
0Contents lists available at ScienceDirect
Case Studies in Thermal Engineering
ELSEVIER journal homepage: www.elsevier.com/locate/csite
Thermal heat transfer fluid problems following a system flush with caustic and water
Christopher Ian Wright *
Global Group, Cold Meece Estate, Cold Meece, Staffordshire ST15 0SP, United Kingdom
ARTICLE INFO ABSTRACT
Heat transfer fluid (HTF) ageing is a complicated chemical process. Laboratory techniques can provide rapid insights into the status of a HTF and a HTF system.
In the current case, a potential client had requested their newly charged HTF be analysed. Prior to filling, however, the system had been flushed with caustic and water. The client reported reduced flow rates, high sludge formation in filters and regular HTF top-ups.
Laboratory testing indicated that the HTF was showing signs of serious thermal cracking (high carbon levels and low flash point temperatures) and significant thermal oxidation (a high total acid number). The recommendation was to drain the HTF from the system and flush the system to remove carbon, acids and flammable by-products. This action would work to reduce the risks associated with coke depositing on the internal pipework of the system and eliminate any fire risk presented by the formation of flammable by-products within the system.
The case highlights the detrimental effects of HTF decomposition on a system as well as the need to flush a system with a fluid intended to be used as a flush and to washout any residual cleaner prior to filling with a new HTF.
© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC
BY license (http://creativecommons.org/licenses/by/3.0/).
Article history:
Received 13 November 2013
Received in revised form
15 January 2014
Accepted 16 January 2014
Available online 24 January 2014
Keywords: Laboratory analysis Thermal fluid Heat transfer fluid
1. Introduction
The ageing of a heat transfer fluid (HTF) is a complicated chemical process [1] and the sporadic analysis of fluids, such as measurement of carbon residue, gives critical insights into the ageing process. Factors that influence the rate of ageing and the longevity of a HTF include elevations in heat, oxidation and materials in contact with the fluid such as pipework and particulates circulating within the fluid itself. The current case describes a particular industrial application where the ageing of the HTF was accelerated. The consequential effects are discussed in the context of the parameters analysed.
In March 2012 a new client, based in South America and operating in the area of textile manufacture, contacted the Global Group to advice on the condition of their HTF following a flush with caustic and then water. The system involved held approximately 4000 l of HTF and ran at 230 °C 24-h per day. After only 6 months, however, the client was physically removing carbon-sludge from their filters every 2-3 days and they had to top-up their system with a further 400 l of HTF. In addition, the expansion tank was running at in excess of 60 °C [1] and was operating without a nitrogen blanket.
Abbreviations: HTF, heat transfer fluid; TAN, total acid number; KOH, potassium hydroxide
* Tel.: +44 7967 230 155. E-mail address: chrisw@globalgroup.org
http://dx.doi.org/10.1016/j.csite.2014.01.003
2214-157X © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).
Results following testing of the client's mineral-based HTF.
Parameters analysed Test method Typical values Analysis results Rating
Carbon residue (%, weight) 1P14 0.02 3.46
Total acid number (mg KOH/g) 1P139 <0.05 0.57
Closed flash point temperature (°C) ASTM D93 210 120
Open flash point temperature (°C) ASTM D92 260 146
Fire point temperature (°C) ASTM D92 Not reporteda 180
Water content (ppm) ASTM D6304 46 45
Ferrous debris (insols) PQ Analex method <10 34
Elements (Fe, Si) ASTM D5185 0 3
Note: A red, amber and green rating system was utilised to indicate the status of the HTF. Red indicating the most severe rating and green the least severe. Colour coding is as follows: green, satisfactory; yellow, caution, amber, action; and red, serious rating. KOH, potassium hydroxide; Fe, iron; Si, silicon. a Not reported in the manufacturer's material safety data sheet.
2. Experimental methods
In order to assess the condition of the HTF and the system, a sample of the fluid was analysed and the results are given in Table 1.
3. Engineering analysis
The parameters analysed in Table 1 are discussed in further detail below.
3.1. Appearance, water, ferrous debris and elements
The appearance of the HTF was described as containing solid debris. Analysis results showed that water, ferrous debris and elements were below levels considered serious or requiring action to remedy the change. In contrast, carbon residue and the open flash point temperature had changed significantly and were at serious levels.
3.2. Carbon residue
Carbon residue was assessed. This measures the unevaporated carbon (coke) residue after a sample of oil is completely evaporated by heating and the oil vapour burned. The carbon residue is weighed and represented as a percentage of the original amount of oil. Carbon residue indicates the formation process whereby high molecular compounds are transforming into coke-like deposits. Hence, it is an indicator of ageing. In this particular case, the percentage carbon residue was nearly 3.5-times the serious level of 1% carbon residue. Open flash point temperature was 146 °C (the level defined as serious being r 150 °C), a drop of 114 °C (- 44% from 260 °C). Assuming that the sample is representative of the HTF in the whole system, the HTF was showing signs of serious thermal cracking.
The elevated level of carbon leads to system fouling in which carbon will deposit and bake on the internal surfaces of system. The window for a change, and action to be taken, is ideally whilst the carbon residue level is less than 1% weight (i.e., between 0.75 and 1.00% weight). This is where the carbon is still soft and can be cleaned off using a thermal cleaning and flushing fluid. However, above the serious level of 1%, carbon starts to bake on to the internal pipework and cannot be easily removed. The consequence of this being that the system starts to lose its thermal efficiency as carbon has an insulating effect and the system will require more energy to heat the HTF. The baking of carbon on to the internal pipework is the most common form of heater/coil failure as hot spots can occur. In such cases, the heater coil breaches and at this stage a thermal HTF fire may occur. One important point to mention here is that systems should be designed with safety interlocks. For example, a "fusible" link to cut-off the gas supply in the event of fire within the heater and pumps that cutout to prevent further fuelling of a fire. Some systems also have an internal fire damping system within the heater. This is the ideal scenario but in practice, these may not be installed or have not worked, which is often due to poor maintenance.
C.I. Wright / Case Studies in Thermal Engineering 2 (2014) 91 -94
3.3. Flash point temperature
The open flash point temperature was assessed and judged to be at a serious level (r 150 °C). Flash point is the temperature at which liquid fuels develop volatile vapours. There are two methods to assess flash point temperature - open and closed. Open flash tests mimic the scenario of vapours mixing with air and being partly removed by air movement as the most volatile escape. This leads to a slightly higher test value because of the air being added. Closed flash point temperature mimics vapours not mixing with air (i.e., being kept together and the most volatile are maintained in the fluid) and so the test value is lower. Flash point temperature reflects the extent of flammable decomposition products. The flash point is the lowest temperature where there is sufficient flammable vapour to ignite in the presence of an ignition source. In this case, the open flash temperature had decreased from 260 to 146 °C (a decreased of 44%). Closed flash point temperature was judged as being cautionary (i.e., between r 130 and > 110 °C) and decreased by 57% from a starting value of 210 °C. The dropping closed flash point temperature indicates that the system is not venting effectively and so highly volatile decomposition products are starting to accumulate.
3.4. Total acid number (TAN)
TAN and fire point were defined as being cautionary. The TAN assesses those components with an acid function (weak organic acids). The test quantifies the amount of potassium hydroxide that is needed to neutralise free acids contained in a 1-mg sample of the test oil. In the form of water solid acids, these can be particularly damaging as they attack the material in contact with the oil and can lead to corrosion of pipework. A high TAN is indicative of ageing of the fluid, oxidation progress and corrosion problems in a system. In this particular case, the TAN was 0.57 and was judged to require action (i.e., between z 0.4 to < 1.0). If left untreated this figure would continue to climb and lead to the exponential formation of heavy ends (more carbon) as well as corrosion, particularly in the expansion tank. From the client's description it was clear that the build-up of sludge was occurring as it was physically being removed from the filter every couple of days. It is likely that TAN would have been higher if this were not being done. This is reflected in Fig. 1 where the percentage carbon residue is plotted against TAN. The equation for the fitted linear line is y=0.448x+0.080 (linear correlation coefficient; R2=0.361) and indicates that TAN is approximately half the value of carbon at any given point on this line. Therefore, using the above equation, at a carbon value of 3.46 the TAN would be estimated to be around 1.55. A serious rating for TAN is z 1.0 mg KOH/g.
It is normally recommended that customers dilute their fluid to achieve a TAN below 0.2 mg KOH/g. At levels below this, the effects of accelerated carbon production and corrosion are reduced to insignificant levels or to values close to those seen with a new HTF.
3.5. Viscosity
A further parameter that was assessed was viscosity at 40 °C, which represents a change in the fluid's molecular sizes and structures. Interestingly, viscosity started at 54 centistokes and decreased to 27.3 centistokes (- 57%). This suggests that the fluid is effectively thinning which ties in with the increased formation of volatile products. It does not tie in, however, with the increasing TAN and carbon value, which should work to increase viscosity. The simple explanation for this is that the sludge and carbon being formed was being removed from the system through filtration. Therefore, it was not leading to thickening of the fluid as might be expected if the oxidation products remained in solution [1]. It is also likely that not all the
Fig. 1. Plots of carbon residue versus total acid number (TAN). Note: Paired data extrapolated from the test results of four of our client's systems that are using a mineral based thermal fluid. 86 paired data points are shown from test results recorded over the last 5 years whilst these systems were sampled. The fluid being used in the above plot was Globaltherm® M (a mineral based HTF).
carbon is being removed from the system and was actually adhering to the internal surfaces. Indeed, the client did indicate that the thermal fluid, when new, was flowing at a rate of 22 m per minute but when sampled, its flow rate was 14 m per minute (a reduction of 36.4% over roughly a 6 month period).
4. Conclusions and recommendations
4.1. System findings
The test results from the oil sample indicate both the health status of the system and the fluid within the system. The system was in a dire state owing to the build-up of carbon and formation of coke deposits on the internal pipework of the system. The system was also at risk owing to the formation of flammable by-products.
4.2. Heat transfer fluid findings
From a HTF perspective, the fluid was showing signs of serious thermal cracking and significant thermal oxidation. Indeed, a TAN of 0.54 is significant when taken in together with the accelerated carbon production and these would explain why carbon residue was excess of 3%. The seriousness of thermal cracking is also reflected by the low flash point temperature (i.e., increased composition of highly volatile decomposition products).
4.3. Recommendations
The immediate actions recommended were to drain the system of its current HTF and to replace it following cleaning and flushing of the system to remove carbon, acids and flammable by-products. Flushing would also work to remove carbon deposits and solid debris, which could potentially damage pump seals whilst in circulation. However, flushing does not remove carbon that is baked on to the internal pipework. In this case, a more aggressive cleaner would be required. It is important to remember that any cleaner would need to be removed from the system and a second flush would be recommended to ensure the cleaner was not left to reactive with the new HTF. Something that the client had not considered as they had cleaned the system with caustic and water, but it would seem that these were not flushed from the system prior to charging the system with a new HTF.
4.4. Importance of flushing and maintenance
This case highlights that thermal HTF maintenance should remain at the forefront of all engineers and maintenance managers so that a safe and an efficient operation is always maintained. This case also shows the detrimental effects of a decomposing HTF on a system's pipework. It also emphasises the need to flush a system with a flushing fluid and to washout any residual cleaner prior to filling with a new HTF. This is extremely important, as HTF's can be expensive to replace (anywhere up to $20 per kg depending on HTF type and volume required). Regular HTF maintenance works to prolong the life of the HTF and is much cheaper than replacing the HTF or the HTF system. Indeed, the last recommendation is that engineers and maintenance managers should consider implementing a regular sampling regime to monitor the health of their HTF.
References
[1] Wagner O Walter. Heat transfer technique with organic media. In: Heat transfer media, second ed. Graefelfing, Germany: Maria-Eich-Strape; 1997. p. 4-58 [Chapter 2].
