CC BY
0CASE STUDIES IN THERMAL ENGINEERING
Author's Accepted Manuscript
Heat transfer system safety: Comparing the effectiveness of batch venting and a lightends removal kit (LERK)
Christopher Ian Wright, Julien Premel
www.elsevier.com/locate/csite
PII: S2214-157X(14)00029-X
DOI: http://dx.doi.org/10.1016/j.csite.2014.09.001
Reference: CSITE46
To appear in: Case Studies in Thermal Engineering
Received date: 11 August 2014
Revised date: 3 September 2014
Accepted date: 5 September 2014
Cite this article as: Christopher Ian Wright, Julien Premel, Heat transfer system safety: Comparing the effectiveness of batch venting and a light-ends removal kit (LERK), Case Studies in Thermal Engineering , http://dx.doi.org/10.1016/j. csite.2014.09.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Title: Heat transfer system safety: Comparing the effectiveness of batch venting and a light-ends removal kit (LERK).
Abbreviated title: Batch venting
Author: Christopher Ian Wright, BSc, PhD, MBA; 2Julien Premel. Corresponding author: Christopher Ian Wright Corresponding e-mail: chrisw@globalgroup.org Corresponding telephone: +44-7967-230-155 Authors contact addresses:
1Global Group; Cold Meece Estate, Cold Meece, Staffordshire, ST15 0SP, UK.
2SNCF; Technicentre Paris Rive Gauche, 91-103 Avenue Marx Dormoy, 92220 Bagneux,
France,
Abstract (word count = 200)
The heat transfer fluids (HTF) should be analysed at least once per year to determine the extent of thermal degradation. Under normal operating conditions, mineral-based HTFs will thermally degrade and so the bonds between hydrocarbons break to from shorter-chain hydrocarbons known as "light-ends." These light-ends accumulate in a HTF system and present a potential fire risk. Light-ends can be removed from a HTF system via a batch vent or installation of a temporary permanently installed light-ends removal kit (LERK). Data was collected prior to and following batch venting or installation of a LERK. The main study parameter was closed flash temperature as open flash temperature and fire point did not change considerably. Analysis showed that both methods increased closed flash temperature in excess of 1300C three months after the intervention, so both methods were deemed effective. Data showed that the percentage change achieved with the LERK, compared to batch venting, was 2-fold higher at three months and 10-fold higher at 6 months. The duration of effect was longer with the LERK with closed flash temperature being stable and consistently above 1300C for 50 months after being permanently installed. This case highlights the effectiveness of a permanently fitted LERK which is effective for the longer-term control of closed flash temperature. However, mobile LERKs could be an option for manufacturers looking to manage closed flash temperature on a shorter-term basis or as an alternative to batch venting.
Key words
Light-ends; closed flash point temperature; batch venting; light-ends removal kit; fire risk.
Abbreviations
HTF, heat transfer fluid; LERK, light-ends removal kit.
Introduction
In the UK there is estimated to be 4,000 heat transfer systems using a heat transfer fluid (HTF; e.g., water, mineral-based HTFs) to transfer heat energy to process equipment [1]. The continued operation of mineral-based HTF systems depends on this HTF being sustained for as long as possible as they thermally degrade over time. For this very reason, manufacturers recommend that a HTF is sampled at least once annually if a HTF operates close to its upper operating temperature or every other year if it is operating more than 20OC below its upper operating temperature [2]. Some insurers also stipulate annual HTF sampling [3]. HTF sampling and analysis enables the assessment of oxidative state and to determine the presence of foreign contaminants such as water [4, 5]. The formation of short-chained hydrocarbons or "light-ends" is a by-product of thermal degradation and form when long-chained hydrocarbons start to break and form shorter-chained light-ends.
Over time all HTFs will thermally degrade when operating at high temperature and are likely to lead to losses in efficiency, potential pump failure, an increased flammability risk all of which may cause system downtime and increased costs [6]. Light-ends are one aspect of HTF degradation and can be monitored by routine laboratory testing of open and closed flash temperatures [4, 5]. These parameters decrease as the composition of light-ends in a HTF increases. Flash temperature represents the proportion of flammable decomposition products or light-ends in a HTF and an increase in light-ends is denoted by a decrease in flash temperature. A decrease in flash temperature represents an increasing potential fire risk [5]. This is because light-ends have lower boiling and ignition temperatures and these represent a potential risk for system fires. Testing uses a flame and exposes the HTF to flame to gain a temperature at which the HTF flashes [5]. This is conducted in the laboratory according to internationally recognised testing standards (i.e., ASTM D92 and D93).
Any HTF system in Europe that operates above the flash temperature of the HTF will need to manage the potential fire risk in-line with ATEX Directive 99/92/EC (ATEX 137 or ATEX Workplace Directive) and ATEX Directive 94/9/EC (ATEX 95 or ATEX Equipment Directive). In the UK Directive 99/92/EC was put into effect through regulations 7 and 11 of the Dangerous Substances and Explosive Atmospheres Regulations 2002 (DSEAR). Maintaining the condition of a HTF falls under these Directives and regulations as HTFs with a low flash temperature are a potential fire risk. In this case the flash temperature of the HTF becomes a critical risk factor with a decreasing flash temperature lowering the temperature that a HTF will flash and hence the fire risk increases. Taking measures to maintain flash temperature therefore plays an important role in managing the risk of fire.
Two methods commonly utilised to manage flash temperature are batch venting and installation of a LERK [7]. These methods remove or prevent light-ends building-up in a HTF. Batch venting, which is not suitable for every system, involves heating the expansion tank to raise the temperature of the circulating HTF and to vaporise the light-ends from the HTF system. In contrast, a LERK is a permanent system installation that works by distillation and continuously removes light-ends from the HTF and the HTF system [7].
The objective of this research was to determine the effectiveness of batch venting and a LERK in the management of flash temperatures (reflective of light-ends accumulation) in a HTF. Analysis focused on open and closed flash temperatures and fire temperature. These parameters were recorded prior to and following batch venting and installation of a LERK to assess their effect in terms of magnitude of change and duration of this change.
Experimental methods
Number of systems assessed
Three systems had been batch vented and form the basis of experimental analysis. All three systems contained Globaltherm™ M, which is a mineral-based HTF. A light-end removal kit (LERK) can be used to control light-end formation in a HTF. Data obtained following batch venting was compared with that following the installation of a LERK [7].
From the multiple samples taken it is possible to assess changes against time and to assess changes prior to and following a system intervention to control flash point temperatures using either batch venting or installation of a LERK.
How is the HTF sampled?
All systems were sampled whilst the HTF was operational and this means that the system was in full operation at the time the HTF was extracted. A 500 ml sample of the HTF was taken. This is performed using a custom designed sampling device (please see [5, 7]) that isolates the sampled HTF so it is not exposed to air. This stops light-ends vaporising from the HTF and ensures that a true representative sample of the HTF is obtained.
What HTF parameters are analysed in the laboratory?
Once the sampled HTF has cooled to ambient temperature, the characteristics of the HTF are analysed in the laboratory according to IS014001 and IS017025 [7]. The characteristics analysed in this case are summarised in Table 1.
Table 1. Parameters and test methods for flash and fire temperatures.
Characteristic Test method
Fire temperature (°C) ASTM D92 [8]
Open flash temperature (°C) ASTM D92 [8]
Closed flash temperature (°C) ASTM D93 [9]
Note: Appearance, carbon residue, total and si trong acid number, kinematic
viscosity, water content, ferrous wear debris and elements were also analysed but are not the focus of this case report.
How the condition of a mineral-based HTF is rated
The parameters in Table 1 are assigned one of four ratings based on pre-defined criteria. These ratings are outlined in Table 2.
Characteristic Condition rating for mineral-based HTFs
Satisfactory Caution Action Serious
Fire temperature (°C) <255 to >210 <210 to >180 <180 to >150 <150
Open flash temperature (°C) <221 to >160 <160 to >150 <150
Closed flash temperature (°C) <210 to >130 <130 to >110 <110 to >85 <85
Equipment for managing decreases in HTF flash temperatures
i. Batch venting
It is important to mention that batch venting is not suitable for every system and that the expansion tank is heat rated to deal with elevations in temperatures. If the customer is unsure, they should obtain advice from an expert.
Batch venting involves heating the expansion tank to raise the temperature of the circulating HTF and to vaporise the light-ends form the HTF system. The volume of HTF that is heated is managed by opening the valves on the purge line into the expansion tank and allowing the HTF to flow into the expansion tank. At elevated temperature, the rate of oxidation is increased in keeping with Arrhenius's rule [10]. To negate this, a Nitrogen lance is inserted into the HTF to provide a carrier gas and to stop oxidation. Changes in oxidative state are reflected by changes in total acid number (TAN), which is routinely analysed (see Table 1). Light-ends escape during the venting process and vapours are collected into a condenser. Batch venting is continued until no further light-ends are collected.
ii. Light-ends removal kit (LERK)
A LERK can be installed as a temporary or permanent fixture to the HTF system. The LERK works by continuously removing light-ends from the HTF system via distillation. In the current case a permanent LERK was installed. The HTF is sprayed into a distiller and this allows the light-ends to evaporate from the HTF. Light-ends collect in a condenser whilst the HTF continues to circulate. Further details regarding the LERK can be found at the following [7].
Engineering analysis
This case study focused on the analysis of flash and fire temperatures. Analysis was conducted on past analysis reports. All data was analysed using Microsoft Office Excel 2007.
Data from individual systems is reported (four in total containing a mineral based heat transfer fluid) as absolute values and grouped data is shown as means or means ± standard deviation (SD) unless stated otherwise. The presentation of data was standardised with the following time points presented: -6 (>-12 to <-6 months), -3 (>-6 to <0 months), 0 (defined as baseline and the time point immediately prior to batch venting ort installation of the LERK), 3 (>0 to <3 months) and 6 months (>3 to <6 months). Time zero is defined as the baseline value. Two time points were taken prior to (i.e., pre) and post batch or installation of the LERK (post).
Results
Data are presented from five systems in which batch venting was conducted to manage flash point temperatures. For comparison one system, in which light-ends were removed using a permanently installed LERK is presented. All systems used a mineral-based heat transfer fluid. The data below is presented pre and post batch venting or LERK installation.
Pre-intervention
Figure 1 and Table 3 summarise data recorded pre batch venting and pre installation of a LERK. Table 3 shows that between time points -6 and 0, fire temperature was relatively stable and increased by +0.8% from 243.80C. During this time, there was a small decrease (2.6% from 203.40C) in open flash temperature. The largest change was seen in closed flash temperature, which decreased by 20.5% from 143.20C.
The above changes were very similar to those recorded prior in the LERK group. Table 3 shows that open flash and fire points were relatively stable (+5.0% and +2.6% change between time -3 and 0). In keeping with the batch venting group, the greatest change was seen in the closed flash temperature which, during this time, decreased by 25.0% from 109.30C.
Table 3. The effect of batch venting and light-ends removal on flash point and fire temperatures.
Closed flash (0C) Open flash (0C) Fire (0C)
Time (month) Phase Batch vent LERK Batch vent LERK Batch vent LERK
-6 Pre 143.2±19.2 109.3 203.4±11.7 191.0 243.8±4.6 234.3
-3 Pre 121.2±13.1 123.5 195.2±14.0 209.5 253.2±18.4 245.5
0 Baseline 113.8±29.6 82.0 198.2±22.8 196.0 245.8±5.3 246.0
3 Post 161.8±21.7 149.3 198.8±15.5 210.0 242.4±8.3 246.7
6 Post 129.2±20.1 174.0 193.4±21.8 208.0 243.6±7.6 244.0
Note: Bate h venting represents the average from 5 systems. LERK data was adapted from
[7] and taken from one system. Pre and post refer to recording taken before and after batch venting or installation of the LERK.
Post-intervention
In keeping with the pre-intervention levels, open flash and fire temperatures were relatively stable between time 0 and +6 months. Indeed, fire temperature decreased by -0.9% (batch
vent) and -0.8% (LERK) at month +6. In both cases, the change between time 0 and +6 months was less than 5%. The changes in open flash temperature were -2.4% (batch vent) and +6.1% (LERK) at month +6. Similar changes were seen at month +3 (+0.3 and +7.1%, respectively). Thus suggesting that both were relatively stable during this period and this can be seen in Figure 1. In contrast, the changes in closed flash temperature were much larger. For the LERK, closed flash temperature increased by +82.1% (149.30C) and +112.2% (174.00C) at month +3 and month +6 and this exceeded the value recorded at month -6 (i.e., 109.20C). Perhaps the most surprising finding is that closed flash temperature was not affected to the same extent by batch venting. Indeed, it increased by +42.2% at month +3 and +13.5% at month +6. Therefore, between month +3 and month +6, the percentage change was -28.7%. At +6 months, the closed flash temperature was lower than at month -6 (i.e., 143.20C) and just above (+15.40C) the value recorded at time 0 (i.e., 113.80C) (see Figure 1).
Viscosity, carbon and total acid number
These parameters are routinely analysed in the laboratory. Results showed that viscosity increased between -6 and 0 months (+7.9%) and increased between 0 and +6 months (+2.8%). This magnitude of effect was similar to those seen with the LERK (-10.1 and -1.6%, respectively). In the batch venting group, the carbon (percentage weight; IP14) and total acid number (mg of potassium hydroxide per gram; IP139) of the HTF remained within specification and was <0.2 throughout.
Factors to consider when using batch vent or LERK to manage a drop in closed flash temperature
The below Table compares and contrasts the key factors when considering the use of batch venting and the installation of a LERK.
Table 4. Factors to consider when using a batch vent or LERK to manage closed flash temperature
Parameter Batch venting LERK
Management of closed flash temperature? Yes. Yes.
Management of open flash temperature? *No. *No.
Management of fire temperature? *No. *No.
Permanent installation? No, an intermittent methodology to manage closed flash temperature. Yes, permanently installed in the HTF system. Although temporary installations are available.
Provides short-term control of closed flash temperature? Yes, batch venting effectively removes closed flash components in the HTF but this was less effective as the LERK at +3 and +6 months. Yes, effectively removes closed flash components in the HTF and was twice as effective as the LERK after +3 months.
Provides longer-term control of closed flash temperature? Yes, batch venting was effective in the shorter-term but would need to be repeated every 3 to 6 months. Yes, having a larger magnitude of effect (10-fold higher after +6 months) than batch venting with a longer, sustained effect on closed flash point lasting, in this case, in excess of 50 months.
Can this be used as a transient intervention to control closed flash temperature? Yes, batch venting is a transient intervention rather that can be repeated but this would need to be factored into a production schedule Yes, mobile LERKs are an alternative to a permanently fitted LERK offering flexibility in terms of installation (i.e., from several days to several months)
Note: *, this is based on the current data set in which open flash and fire temperatures were relatively well controlled. Hence further work is needed to define the effect of batch venting and LERK installation when these parameters are much lower.
Discussion
The current study assessed the effectiveness of batch venting in the management of flash and fire temperatures. This was compared with the effectiveness of flash and fire temperature management following the installation of a LERK. These methods are commonly used to control the build up of light-ends in a HTF and do so by removing the lower flash temperature components from a HTF. This is important for the longer-term safety of thermal HTF systems. The current study demonstrated that both batch venting and LERK are effective in the removal of lower closed flash temperature components; however, their duration of action differed with batch venting only providing short-term effects as compared with the long-term effects achieved with a permanently installed LERK.
There is estimated to be 4,000 HTF systems in the United Kingdom [11]. These HTF systems use either water (steam) or organic fluids, such as a mineral-based HTF. Mineral-based HTFs are used to achieve high operating temperatures. Under such conditions, these HTFs fall under ATEX Directives and DSEAR 2002. This is because high temperature HTF systems quite commonly operate at temperatures above the flash temperature of the HTF. This means that the ignition temperature is below the operating temperature and this is a potential risk should there be a leak and vapours allowed to form. Many companies are not aware that under high temperature conditions the thermal degradation of mineral-based HTFs can be accelerated and flash temperatures can drop dramatically. In addition, if the HTF becomes contaminated, this process can be further accelerated [4]. Under such circumstances, HTFs operating under normal conditions present a potential for fire and the employer needs to manage the safety of the site and the site workers in line with current Directives and Regulations.
Routine sampling and analysis of HTFs, at least once per year according to the HTF manufacturer [2, 3], is recommended to monitor the condition of a HTF and the potential fire risk in the event of a fluid leak and vapour formation. Sampling and analysis, however, only provide insight into the condition of a HTF. In addition to this, the company needs to ensure they are proactively managing the removal of lower flash point components from a HTF. This can be achieved by batch venting or installation of a LERK [7]. The current study assessed the effectiveness of batch venting in the control of flash and fire temperatures. Open flash temperature and fire point were not changed in the current analysis. These values were judged to be satisfactory (i.e., >160 and >210°C, respectively) and so would not be expected to change considerably as they are close to typical values for a virgin HTF (see Table 2). Closed flash temperature was dramatically lower and on average was rated as cautionary (see Table 2). Analysis showed that batch venting removed closed flash components as
demonstrated by the increase in closed flash temperature at 3 months (+42.2%). This effect was not sustained and at 6 months it had decreased with closed flash temperature only being +13.5% above baseline values. This finding demonstrates the short-term effectiveness of batch venting. Thus highlighting its use in the short-term control of closed flash temperature components.
In keeping with the findings for batch venting, the installation of the LERK had little effect on open flash and fire temperatures, but did had a marked effect on closed flash temperature (see Table 3). The key differences versus batch venting, however, are the magnitude of the effect and the duration of this effect. Indeed, the percentage change at +3 months was double that achieved with batch venting (+82.1 versus +42.2%), although the absolute change at this time point was similar (+48.0 versus 67.20C). At +6 months this difference was even more marked (percentage change, +112.2 versus +13.5%; and, absolute change, +92.0 versus 15.40C). This demonstrates that both batch venting and the installation of a LERK are effective in the management of closed flash temperature. It is important to mention that our past research has shown that a LERK helps to restore closed flash point temperature to stable levels that are similar to those seen for a virgin HTF [7]. This reduces the ongoing HTF maintenance, at least in terms of flash temperatures.
Any company installing a LERK will need to conduct a cost-to-risk-benefit analysis. In terms of cost, the installation of a LERK is roughly 10-times that of a single batch vent. In terms of the risk, flash temperatures need to be managed to ensure site and staff safety. Table 2 uses a sliding scale to define HTF safety with the risk from decreasing closed flash temperature needing to be managed once it drops to <1300C (see Table 2). In the current example, both methods increased closed flash point above this level and into the satisfactory range at +3 months. Lastly, the benefit gained needs discussing in further detail. This has two elements, the magnitude of effect and the duration of effect. Data showed that the magnitude of the LERK's effect on percentage change in closed flash temperature was 2 and 10-fold higher (at +3 and +6 months) as compared to batch venting. Furthermore, the duration of effect was greater with the LERK installed. Indeed, the LERK, in the current example, was installed in January 2010 and the HTF was last analysed in March 2014. At which point closed flash temperature was 1680C. This LERK had been in situ for 50 months and the HTF was judged stable, carrying a satisfactory rating (see Table 2). The batch venting data showed that its effect was sustained for only 3 months as at 6 months closed flash temperature had dropped below 1300C. Therefore, the cost of one LERK is equal to ten batch vents and the LERK would need to be installed for 30 months to gain a return on investment. However, this does ignore the fact that in the case of the batch vent, the closed
flash temperature will continue to go through a cycle of peaks and troughs and closed flash temperature will need to be closely managed as it decreases below actionable levels (<1300C). A further point is that if batch venting is not effective in the management of closed flash temperature, it is likely that some interventions, such as a complete HTF system flush and fill, will be far more costly in the shorter-term.
The cost-to-risk-benefit shows, therefore, that installing a LERK is a more expensive in the shorter-term. However, this cost needs to be considered in terms of the duration of effect. The LERK provided much larger (ten-fold higher at +6 months) and longer effects (50 versus 3 months; at least 17-times longer) than batch venting in the current case. Thus, over the longer-term, the LERK becomes more cost effective. In light of the risk-benefit analysis, data clearly showed that the permanently fitted LERK helped to better manage closed flash temperature over the longer-term with values being sustained above 1300C. The effect on closed flash temperature was larger over the longer-term and lasted longer than that achieved with batch venting. However, the choice to perform a batch vent or install a LERK is driven by the manufacturer. In certain cases, factors may mean that batch veting is preferable to a permanently installed LERK. In this specific case a mobile LERKs could be an option as it offers a flexible solution (i.e., can be installed for a number of days, weeks or months) for manufacturers.
Conclusions and recommendations
Batch venting and the installation of a LERK were shown to be effective in the management of closed flash temperature. The decision on which methodology has to be based on discussions with the client and this needs to take into consideration the cost but also the relative risk-to-benefit. The risk being the safety of the site and the site workers and the benefit being the effective management of closed flash temperature, which needs to be considered in terms of the magnitude and duration of effect. Table 4 compares the key factors that need to be considered when considering a batch vent or installation of a LERK. The key factor is that batch venting reduces the closed flash components in a HTF, but this methodology would potentially need repeating over the longer-term. The installation of a permanently fitted LERK would be more appropriate to the longer-term control of closed flash temperature. However, mobile LERKs could be an option for manufacturers looking to manage closed flash temperature on a shorter-term basis.
Acknowledgements
The authors would like to thank the Global Heat Transfer team: engineers, Danny Bradford, Dave Dyer and Martyn Tinsley; and, technical expertise, Andy Burns and Lisa Cho.
References
[1] D.J. Kukulka, M. Devgun. Fluid temperature and velocity effect on fouling. Applied Thermal Engineering 27 (2007) 2732-2744.
[2] Solutia. Liquid phase systems design guide: A design, operating low-pressure heat guide for low-cost, and maintenance transfer systems. Publication number 7239128D. Source: www.therminol.com/pages/tools/therminol_liquid_phase.pdf. Accessed: 24th April 2014.
[3] Heat transfer by organic and synthetic fluids, in: Factory Mutual 7-99. Property Loss Prevention Data Sheets 12-19. Source: ftp://cable-129-140-
83.b2b2c.ca/sda1/Basement/Mes%20documents/Reconnaissance%202011/Codes_Normes /Factory%20Mutual/DS/7-99.PDF. Accessed: 24th April 2014.
[4] C.I. Wright. Thermal heat transfer fluid problems following a system flush with caustic and water. Case Studies in Thermal Engineering 2 (2014) 91-94.
[5] C.I. Wright, A. Burns, C. Jones. Sampling hot heat transfer fluids: Simple Insights for Gaining a Representative Sample International Journal of Engineering and Innovative Technology 3 (2013) 202-204.
[6] J. McKenna. How safe are heat transfer systems? Strategies for process control and plant management in Process Engineering (2010). Source: www.processengineering.theengineer.co.uk. Accessed: 31st July 2014.
[7] C.I. Wright. Effective management of heat transfer fluid flash point temperatures using a light-ends removal kit (LERK). Case Studies in Thermal Engineering 4 (2014) 9-14.
[8]. ASTM Standards D92 Standard Test Method for Flash and Fire Points by Cleveland Open Cup Tester. Link: http://www.astm.org/Standards/D92.htm. Accessed: 3rd September 2014.
[9] ASTM Standards D93 Test Methods for Flash Point by Pensky-Martens Closed Cup Tester. Link: http://www.astm.org/Standards/D93.htm. Accessed: 3rd September 2014.
[10] The Open University. Engineering: The challenge of temperature. Source: www.open.edu. Accessed: 31st July 2014.
[11] A. McKay, R. Franklin. Good practice: Fire and explosion hazards with thermal fluid systems. Loss Prevention Bulletin 219 (2011) 13-17.
FIGURES
Figure 1. Plots showing closed flash temperature (1A), open flash temperature (1B) and fire temperature (1C) against time following batch venting (open circles) and following installation of a light-ends removal kit (filled circles).
Figure 1 A. Closed flash temperature.
Figure 1 B. Open flash temperature.
Time (month)
Figure 1 C. Fire temperature.
(Li O
f3 1— OJ CL E a; 2^0 c « # # • 0 •
3 -6 -3 0 Time (month) 3 6
Note: Time points plotted are as follows: -6 (>-12 to <-6 months), -3 (>-6 to <0 months), 0 (defined as baseline and the time point immediately prior to batch venting), 3 (>0 to <3 months) and 6 months (>3 to <6 months). Filled circles, LERK; and, open circles, batch venting.
