Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 1 Indroduction Bioenergy will play a key role in achieving the mandatory EU target of 20 % renewable energy in 2020 set by the 31 Aus Wissenschaft und Forschung * Dipl. Ing. (FH) Christoph Schneidhofer Dipl. Ing. Manuel Dorfmeister, BSc. Dr. Nicole Dörr AC2T research GmbH, Wiener Neustadt, Austria In-line corrosion sensor for oil condition monitoring of biogas operated stationary engines C. Schneidhofer, M. Dorfmeister, N. Dörr* Eingereicht: 20. 6. 2015 Nach Begutachtung angenommen: 20. 8. 2015 Die Versäuerung ist ein wichtiger Parameter bei der Bewertung des Ölzustands. Vor allem beim Betrieb von stationären Biogasmotoren ist die Versäuerung einer der Hauptindikatoren für die Veranlassung eines Ölwechsels. Neben einer beschleunigten Ölalterung können die gebildeten Säuren Korrosion an Motorteilen, z. B. Lager, hervorrufen. Demzufolge ist die Online-Überwachung der Versäuerung bzw. Korrosivität des Motoröls mit einem chemisch wirkenden Korrosionssensor zweckmäßig. Kernstück dieses Sensors ist eine Opferschicht aus einem definierten Material, welches vom Öl angegriffen wird, sobald dieses eine korrosive Wirkung gegenüber der Opferschicht entwickelt. Der Materialverlust infolge Korrosion wird mittels kapazitiv gekoppelter Ausleseelektroden laufend verfolgt. Das grundsätzliche Verhalten des Sensors wird anhand statischer Versuche in ausgewählten Modellölen mit typischen sauren Verbindungen diskutiert. Die Online-Anwendbarkeit wurde untersucht, indem der Korrosionssensor in einer Vorrichtung zur künstlichen Alterung von Ölen integriert wurde. Öldaten aus konventionellen Analysen von regelmäßig entnommenen Ölproben konnten mit den Änderungen des Sensorsignals korreliert werden. Ergebnisse aus Feldversuchen mit stationären Gasmotoren unterstreichen die Verwendbarkeit des Sensorkonzeptes. Sowohl bei der künstlichen Alterung als auch bei den Feldversuchen konnte gezeigt werden, dass der Beginn der Korrosion der Opferschicht mit einem kritischen Versäuerungsgrad korreliert und somit für die Anzeige eines Ölwechsels eingesetzt werden kann. Schlüsselwörter Korrosionssensor, Versäuerung, Ölzustandsüberwachung, Ölsensoren, künstliche Alterung, Gasmotor, Großmotor Acidification is a crucial parameter for the evaluation of the condition of lubricating oils, in particular in stationary engines driven by gaseous biofuels. Here, acidification is one of the main indicators for an oil change. Besides the effect of acceleration of oil degradation, acids formed can also cause corrosion on engine parts, e.g. bearings. For this reason, the online monitoring of the engine oil corrosiveness by means of a chemical corrosion sensor is useful. Core piece is the sacrificial layer prepared from a specific material, which will be attacked by the oil engine as soon as it develops substantial corrosiveness against the used material. The material loss due to corrosion is continuously measured by capacitive coupling of the readout electrodes. The fundamental response of the sensor signals to the respective oils was evaluated by static experiments in selected model oils containing typical acidic compounds. The online applicability was investigated by integrating the sensor system in an artificial alteration device. Thereby, laboratory analyses of periodically taken oil samples could be correlated to the changes in the sensor signals. The usefulness of the proposed sensor concept is also illustrated by results from sensor systems installed in stationary gas engines. In both laboratory artificial alteration and field experiments, the sensor signals showed that the initial time of corrosion can be correlated to a critical level of acidification and hence can be applied for the indication of an oil change. Keywords Corrosion sensor, acidification, oil condition monitoring, artificial alteration, gas engine, oil sensors, large engines Kurzfassung Abstract T+S_3_16 05.04.16 09: 01 Seite 31 32 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 renewable energy directive. Bioenergy currently covers more than 10 % of the final energy demand in Europe [1]. Biomass refers to renewable energy coming from biological material such as manure, organic waste, and plants. It is the largest and most important renewable energy option at present. The biogas sector has seen a rapid development in recent years and is continuing to grow. However, biogas engine operators suffer from high costs associated with maintenance due to the special chemical attack of the engine and engine oil stressed by the buildup of aggressive compounds formed by biogas combustion. Biofuels and their combustion products usually differ significantly in chemistry from fossil fuels resulting in interactions with oil additives that result in accelerated oil degradation and hence significantly shorter oil intervals [2]. Compared to conventional fuels, bio-fuelled engines are more severely exposed to an acidification of the lubricating oil and humidity increase. This results from the fact that biofuels often contain significantly higher amounts of contaminants, in particular sulphur compounds such as hydrogen sulphide, which are oxidised to give aggressive acidic compounds during combustion, thus contaminate the oil, and increase the danger of corrosion of engine parts [3, 4]. Therefore, the corrosion sensor [5-7] as part of a multiparameter oil condition monitoring system is proposed, able to measure the direct impact of the oil corrosiveness. Core piece of this sensor concept is a sacrificial layer prepared from a material which is exposed to the lubricant in application. Consequently, the sacrificial layer is also corroded once the lubricant develops significant acidification or corrosiveness, respectively. The material loss is monitored in form of the remaining area using capacitively coupled electrodes. In this contribution, the fundamental behaviour of the corrosion sensor based on experiments under static conditions using defined model oils containing base oil, acidic components and selected additives is reported. Furthermore, the sensor was evaluated in so-called artificial alteration experiments simulating the impact of biogas to engine oils for the lab-based evaluation of the online applicability of the proposed sensor concept. Results from the field evidence the usefulness of the proposed sensor. 2 Experimental setup 2.1 Corrosion sensor design Figure 1 shows the schematic setup of the corrosion sensor [7]. Main part of the proposed sensor is the sensing element that consists of a metal film - the sacrificial layer - deposited on an inert substrate. For these investigations, lead was applied with a thickness of 600 nm as material for the sacrificial layer. After fabrication, the metal films of the sensor were covered with a protective lacquer to avoid an undesired surface passivation by air during storage prior to use. For the implementation in corrosion experiments, a defined setup was fabricated consisting of the sensor body including the electrodes for the readout and a sensor cap for fixing the sacrificial element to the sensor body. An appropriate electronic measurement circuit which was directly connected to the readout electrodes evaluated the resonance frequency of the capacitive coupling between the electrodes and the metal film. Therefore, the output frequency is a parameter for the remaining area of the sacrificial layer. As parameter for the corrosiveness of the lubricant monitored, the initial time, i. e. onset of corrosion can be used representing the time when the sensor signal indicates that the remaining area of the sacrificial layers begins to decrease. 2.2 Static laboratory experiments For corrosion experiments under static conditions, substrates carrying the sacrificial layer were cleaned by a solvent to remove the protective layer. Afterwards, the dried substrate was mounted on the sensor base body and fixed with the sensor cap as shown in Figure 1. After these preparatory steps, the fully assembled sensor was immediately immersed into 40 mL of the oil sample in a hermetically Aus Wissenschaft und Forschung results from sensor systems installed in stationary gas engines. In both laboratory artificial alteration and field experiments, the sensor signals showed that the initial time of corrosion can be correlated to a critical level of acidification and hence can be applied for the indication of an oil change. Keywords Corrosion sensor, acidification, oil condition monitoring, artificial alteration, gas engine, oil sensors, large engines 1. Indroduction Bioenergy will play a key role in achieving the mandatory EU target of 20 % renewable energy in 2020, set by the renewable energy directive. Bioenergy currently covers more than 10 % of the final energy demand in Europe [1]. Biomass refers to renewable energy coming from biological material such as manure, organic waste, and plants. It is the largest and most important renewable energy option at present. The biogas sector has seen a rapid development in recent years and continues to grow. However, biogas engine operators suffer from high costs associated with maintenance due to the special chemical attack of the engine and engine oil stressed by the buildup of aggressive compounds formed by biogas combustion. Biofuels and their combustion products usually differ significantly in chemistry from fossil fuels resulting in interactions with oil additives that result in accelerated oil degradation and hence significantly shorter oil intervals [2]. Compared to conventional fuels, bio-fuelled engines are more severely exposed to an acidification of the lubricating oil and humidity increase. This results from the fact that biofuels often contain significantly higher amounts of contaminants, in particular sulfur compounds such as hydrogen sulfide, which are oxidised to give aggressive acidic compounds during combustion, thus contaminate the oil, and increase the danger of corrosion of engine parts [3, 4]. Therefore, the corrosion sensor [5-7] as part of a multiparameter oil condition monitoring system is proposed, able to measure the direct impact of the oil corrosiveness. Core piece of this sensor concept is a sacrificial layer prepared from a material which is exposed to the lubricant in application. Consequently, the sacrificial layer is also corroded once the lubricant develops significant acidification or corrosiveness, respectively. The material loss is monitored in form of the remaining area using capacitively coupled electrodes. In this contribution, the fundamental behaviour of the corrosion sensor based on experiments under static conditions using defined model oils containing base oil, acidic components and selected additives is reported. Furthermore, the sensor was evaluated in so-called artificial alteration experiments simulating the impact of biogas to engine oils for the lab-based evaluation of the online applicability of the proposed sensor concept. Results from the field evidence the usefulness of the proposed sensor. 2. Experimental setup 2.1. Corrosion sensor design Figure 1 shows the schematic setup of the corrosion sensor [7]. Main part of the proposed sensor is the sensing element that consists of a metal film - the sacrificial layer - deposited on an inert substrate. For these investigations, lead was applied with a thickness of 600 nm as material for the sacrificial layer. After fabrication, the metal films of the sensor were covered with a protective lacquer to avoid an undesired surface passivation by air during storage prior to use. metal film oil substrate readout electrodes sensor body sensor cap Figure 1: Schematic setup of the corrosion sensor modified from [7] For the implementation in corrosion experiments, a defined setup was fabricated consisting of the sensor body including the electrodes for the readout and a sensor cap for fixing the sacrificial element to the sensor body. An appropriate electronic measurement circuit which was directly connected to the readout electrodes evaluated the resonance frequency of the capacitive coupling Figure 1: Schematic setup of the corrosion sensor modified from [7] T+S_3_16 05.04.16 09: 01 Seite 32 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 sealed vessel. In such a vessel, a further chemical alteration of the oil was prevented and hence allowed constant, i. e. static, conditions. The assembled test units were kept at 120 °C to monitor the corrosion process at an elevated rate. The chosen elevated temperature accelerated the corrosion process due to a higher rate of the chemical reactions initiated. Under these conditions, the corrosion process of the sacrificial layer mainly depends on the current corrosiveness of the oil sample: the sacrificial layer corrodes faster in more corrosive oils. In the event that no corrosion of the sacrificial layer could be observed after 120 hours, the experiment was aborted and consequently the oil investigated was classified as non-corrosive. The behaviour of the corrosion sensor was determined with a series of defined model oils containing acidic compounds and selected additives. Table 1 lists the oil samples used for these investigations. form about the oil performance. Most standards examine the thermal-oxidative stability of oils by exposing the lubricant to elevated temperatures and oxygen from air e. g. ASTM D4871 [8] or DIN 51352 [9]. Based on the existing methods, an adapted alteration method was used to simulate oil life cycles in gas engines. Therefore, the lubricant was deteriorated at elevated temperature - in this case 160 °C - in a defined vessel able to blow gas into the oil. Additionally, a fixation was used to implement the corrosion sensor setup in the artificial alteration device. In order to simulate the impact of different fuel qualities on lubricant degradation, two different alteration methods were applied. In the first method, air was blown through the lubricant at a rate of 10 L/ h. This method denoted with “air” simulates mild conditions in an engine, i. e. the operation with natural gas which is almost free from contaminants forming aggressive compounds. The conditions for the second method called “biogas” were based on the “air” method and adjusted to simulate the impact of biogas causing accelerated oil degradation. A commonly available gas engine oil was artificially altered under these conditions for around 120 h applying both methods. For monitoring the trend of oil degradation during the alteration procedure, oil samples were periodically taken and analysed in the laboratory. Among others, total base number (TBN) according to DIN ISO 3771 [10] and acid number (AN) according to DIN 51558-1 [11] were determined. TBN indicates the amount of bases, i. e. base reserve, in an oil possible to neutralise acidic components taken up or formed upon lubricant degradation whereas AN refers to the amount of acidic components in an oil. 2.4 Field test In order to evaluate the corrosion sensor under real conditions, the sensor was integrated in the oil circulation system of a stationary gas engine driven with biogas from a wood gasification facility [12]. The sensor signals were recorded every minute. The operating temperature of the engine oil at the sensor position was approximately 80 °C. During the field test, two complete oil life cycles were monitored including the sensor. Furthermore, the gas engine was operated with two different oil types. The gas engine oil used in the first cycle was also selected for artificial alteration in the laboratory. During the observation period, oil samples were taken every 250 operating hours (OH) for the first cycle and every 100 OH for the second, respectively, to carry out reference measurements in the laboratory. 33 Aus Wissenschaft und Forschung Table 1: Model oils for static experiments sample no. oil type base oil + M1 1 mg/ g long chain length organic acid base oil + M2 2.0 % detergent + 1 mg/ g long chain length organic acid base oil + M3 2.0 % detergent + 5 mg/ g long chain length organic acid base oil + M4 0.5 % antioxidant + 1 mg/ g long chain length organic acid 2.3 Artificial oil alteration The online applicability was found out by the integration of the corrosion sensor in an in-house artificial alteration apparatus. This lab-based apparatus is specifically designed for the simulation of the impact of biogases on lubricants during application. Artificial alteration means the application of elevated stress conditions to the oil sample to accelerate oil degradation in a short-term scale. Nevertheless, an artificial alteration procedure should remain as close to reality as possible. This way, the quality of lubricants under specific operating conditions and the usefulness of sensor systems for particular applications can be investigated without the need to perform time-consuming and expensive field tests. Currently used standardised methods for the artificial alteration of engine oils evaluate various parameters to in- T+S_3_16 05.04.16 09: 01 Seite 33 34 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 3 Results and discussion 3.1 Evaluation of static experiments In static laboratory experiments, the sensors were immersed in defined model oils. These model oils contained acidic compounds as well as additives to evaluate the interaction of additives and acids. The results (Figure 2) from static corrosion experiments clearly showed that the model oil containing antioxidant and acids (M4) corrodes nearly identically than the model oil containing only the acidic component (M1). It can be concluded that the antioxidant does not influence the corrosion process which was only determined by the acidic component. Thus, the antioxidant is not able to prevent the metal film from corrosion and consequently it cannot neutralise the acid in the oil. The experiments with detergent and a low amount of acid (M2) showed no significant corrosive attack to the metal film during the entire experiment. In this case, the detergent was able to neutralise the acidic component and/ or to protect the surface against corrosion. The experiment with large amounts of acid (M3) showed a negligible corrosive attack to the metal film for about 50 hours followed by a slow steady decrease of the remaining area. The observed behaviour showed, that the detergent was not able to neutralise the total amount of acidic component. Therefore, a small quantity of reactive substances left, which induced slight corrosion of the sacrificial layer. 3.2 Results from artificial alteration The trends of laboratory analyses of TBN and AN of oil samples regularly taken during artificial alteration are shown in Figure 3. It can be clearly seen that more pronounced oil degradation took place by using the artificial alteration method “biogas” compared to “air”. Such behaviour in “biogas” is expressed by rapid TBN decrease as well as high increase of AN in comparison to “air”. Conventional limits for gas engine oils exemplarily applied to TBN and AN recommend an oil change when base reserve consumption reaches half of the TBN of the fresh oil and AN increases by 2.5 mg KOH/ g in comparison to the fresh oil, respectively [13]. With “biogas”, these limits were reached after around 30 hours of artificial alteration for TBN and then a few hours later for AN, respectively. In the case of “air”, the limits were not reached till the end of the experiments. By looking at the sensor signals recorded during artificial alteration of the gas engine oil, an obvious shorter initial time of corrosion of about 20 hours with “biogas” can be seen in comparison to “air” causing an initial time of corrosion of about 70 hours (Figure 4). Until the initial time of corrosion, the recorded signal is almost stable showing that the sacrificial layer is not substantially corroded so far. After the initial time of corrosion, the signal for the remaining area decreases rapidly due to corrosive attack of the sacrificial layer of the sensor until the entire metal film is corroded. Aus Wissenschaft und Forschung denoted with “air” simulates mild conditions in an engine, i.e. the operation with natural gas which is almost free from contaminants forming aggressive compounds. The conditions for the second method called “biogas” were based on the “air” method and adjusted to simulate the impact of biogas causing accelerated oil degradation. A commonly available gas engine oil was artificially altered under these conditions for around 120 h applying both methods. For monitoring the trend of oil degradation during the alteration procedure, oil samples were periodically taken and analysed in the laboratory. Among others, total base number (TBN) according to DIN ISO 3771 [10] and acid number (AN) according to DIN 51558-1 [11] were determined. TBN indicates the amount of bases, i.e. base reserve, in an oil possible to neutralise acidic components taken up or formed upon lubricant degradation whereas AN refers to the amount of acidic components in an oil. 2.4. Field test In order to evaluate the corrosion sensor under real conditions, the sensor was integrated in the oil circulation system of a stationary gas engine driven with biogas from a wood gasification facility [12]. The sensor signals were recorded every minute. The operating temperature of the engine oil at the sensor position was approximately 80 °C. During the field test, two complete oil life cycles were monitored including the sensor. Furthermore, the gas engine was operated with two different oil types. The gas engine oil used in the first cycle was also selected for artificial alteration in the laboratory. During the observation period, oil samples were taken every 250 operating hours (OH) for the first cycle and every 100 OH for the second, respectively, to carry out reference measurements in the laboratory. 3. Results and discussion 3.1. Evaluation of static experiments In static laboratory experiments, the sensors were immersed in defined model oils. These model oils contained acidic compounds as well as additives to evaluate the interaction of additives and acids. The results (Figure 2) from static corrosion experiments clearly showed that the model oil containing antioxidant and acids (M4) corrodes nearly identically than the model oil containing only the acidic component (M1). It can be concluded that the antioxidant does not influence the corrosion process which was only determined by the acidic component. Thus, the antioxidant is not able to prevent the metal film from corrosion and consequently it cannot neutralise the acid in the oil. 0 20 40 60 80 100 0 20 40 60 80 100 120 140 experimental time [h] sensor signal remaining area [%] M1 M2 M3 M4 Figure 2: Corrosion experiments performed under static conditions with model oils containing additives and acidic compounds The experiments with detergent and a low amount of acid (M2) showed no significant corrosive attack to the metal film during the entire experiment. In this case, the detergent was able to neutralise the acidic component and/ or to protect the surface against corrosion. The experiment with large amounts of acid (M3) showed a negligible corrosive attack to the metal film for about 50 hours followed by a slow steady decrease of the remaining area. The observed behaviour showed that the detergent was not able to neutralise the total amount of acidic component and therefore a small quantity of reactive substances left which induced slight corrosion of the sacrificial layer. 3.2. Results from artificial alteration The trends of laboratory analyses of TBN and AN of oil samples regularly taken during artificial alteration are shown in Figure 3. It can be clearly seen that more pronounced oil degradation took place by using the artificial alteration method “biogas” compared to “air”. Such behaviour in “biogas” is expressed by rapid TBN decrease as well as high increase of AN in comparison to “air”. Conventional limits for gas engine oils exemplarily applied to TBN and AN recommend an oil change when base reserve consumption reaches half of the TBN of the fresh oil and AN increases by 2.5 mg KOH/ g in comparison to the fresh oil, respectively [13]. With “biogas”, these limits were reached after around 30 hours of artificial Figure 2: Corrosion experiments performed under static conditions with model oils containing additives and acidic compounds alteration for TBN and then a few hours later for AN, respectively. In the case of “air”, the limits were not reached till the end of the experiments. 0 1 2 3 4 5 6 7 8 9 10 0 20 40 60 80 100 120 TBN or AN [mg KOH/ g] alteration time [h] TBN biogas AN biogas TBN air AN air Figure 3: Trend of TBN and AN during artificial alteration with “biogas” and “air” method By looking at the sensor signals recorded during artificial alteration of the gas engine oil, an obvious shorter initial time of corrosion of about 20 hours with “biogas” can be seen in comparison to “air” causing an initial time of corrosion of about 70 hours (Figure 4). Until the initial time of corrosion, the recorded signal is almost stable showing that the sacrificial layer is not substantially corroded so far. After the initial time of corrosion, the signal for the remaining area decreases rapidly due to corrosive attack of the sacrificial layer of the sensor until the entire metal film is corroded. 0 20 40 60 80 100 Figure 4: Trend of the sensor signal expressed as remaining area during artificial alteration by “air” and “biogas” method At the initial time of the corrosion with both alteration methods, the oils showed an acidification (AN) of 2 to 2.5 mg KOH/ g and a base reserve (TBN) of about 6 to 7 mg KOH/ g. These results give evidence that the sensor is sensitive to a certain oil condition reached after a defined stress level independent from the time needed to achieve this condition. In other words, the sensor is capable of condition monitoring of engine oils run with different fuel qualities. An additional correlation with data from laboratory analysis is done in chapter 3.4. 3.3. Results from field tests The usefulness of the proposed sensor concept is also illustrated by results determined by online condition monitoring in stationary gas engines. Figure 5 shows the trend of the sensor signal of a sensor system implemented into a gas engine driven with wood gas during two oil life cycles. It is noticed that the trends depicted include only periods with running engine. Nonoperating periods are not considered here. 0 20 40 60 80 100 0 500 1000 1500 2000 2500 3000 3500 Figure 5: Trend of the sensor signal expressed as remaining area during the field tests At the first cycle after about 1300 OH, a significant decrease of the sensor signal due to corrosion of the sacrificial layer was observed. The entire metal film was corroded after about 2200 OH. Before the onset of corrosion after 1300 OH, the sensor showed an almost constant signal, i.e. base signal, as no significant corrosive attack to the sacrificial layer occurred. At the initial time of corrosion, the oil was characterised by an acidification of about 2 mg KOH/ g and a TBN of about 7 mg KOH/ g. This critical oil condition was also detected during artificial alteration (see 3.2.). At the second oil cycle, the initial time of corrosion appeared clearly earlier. Here, the corrosion process started after around 400 OH which is mainly attributed to the fact that another gas engine oil was used in this oil life cycle. According to the results, it is concluded that this oil has a lower quality than that used in the first oil life cycle. Figure 3: Trend of TBN and AN during artificial alteration with “biogas” and “air” method T+S_3_16 05.04.16 09: 01 Seite 34 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 At the initial time of the corrosion with both alteration methods, the oils showed an acidification (AN) of 2 to 2.5 mg KOH/ g and a base reserve (TBN) of about 6 to 7 mg KOH/ g. These results give evidence that the sensor is sensitive to a certain oil condition reached after a defined stress level independent from the time needed to achieve this condition. In other words, the sensor is capable of condition monitoring of engine oils run with different fuel qualities. An additional correlation with data from laboratory analysis is done in chapter 3.4. 3.3 Results from field tests The usefulness of the proposed sensor concept is also illustrated by results determined by online condition monitoring in stationary gas engines. Figure 5 shows the trend of the sensor signal of a sensor system implemented into a gas engine driven with wood gas during two oil life cycles. It is noticed that the trends depicted include only periods with running engine. Non-operating periods are not considered here. At the first cycle after about 1300 OH, a significant decrease of the sensor signal due to corrosion of the sacrificial layer was observed. The entire metal film was corroded after about 2200 OH. Before the onset of corrosion after 1300 OH, the sensor showed an almost constant signal, i. e. base signal, as no significant corrosive attack to the sacrificial layer occurred. At the initial time of corrosion, the oil was characterised by an acidification of about 2 mg KOH/ g and a TBN of about 7 mg KOH/ g. This critical oil condition was also detected during artificial alteration (see 3.2). At the second oil cycle, the initial time of corrosion appeared clearly earlier. Here, the corrosion process started after around 400 OH which is mainly attributed to the fact that another gas engine oil was used in this oil life cycle. According to the results, it is concluded that this oil has a lower quality than that used in the first oil life cycle. 3.4 Sensor signals compared to laboratory analyses The results from the experiments under static conditions (see 3.1) clearly showed that the metal film of the sensor is sensitive to the amount of acidic components in the oil. If there are free reactive molecules which cannot be neutralised by the oil, the sacrificial layer is corroded according to the rule of thumb: the higher the amount of acid the faster the corrosion process. The results obtained from the artificial alteration and field operations suggest that the corrosion process starts at a critical amount of acid in the oil. For the correlation with data from laboratory analysis, the sensor signals at the time of sampling were extracted and compared with the conventional oil parameters AN and TBN determined in the laboratory. As the different types of gas engine oils used showed different initial 35 Aus Wissenschaft und Forschung alteration for TBN and then a few hours later for AN, respectively. In the case of “air”, the limits were not reached till the end of the experiments. Figure 3: Trend of TBN and AN during artificial alteration with “biogas” and “air” method By looking at the sensor signals recorded during artificial alteration of the gas engine oil, an obvious shorter initial time of corrosion of about 20 hours with “biogas” can be seen in comparison to “air” causing an initial time of corrosion of about 70 hours (Figure 4). Until the initial time of corrosion, the recorded signal is almost stable showing that the sacrificial layer is not substantially corroded so far. After the initial time of corrosion, the signal for the remaining area decreases rapidly due to corrosive attack of the sacrificial layer of the sensor until the entire metal film is corroded. 0 20 40 60 80 100 0 20 40 60 80 100 120 alteration time [h] sensor signal remaining area [%] air biogas Figure 4: Trend of the sensor signal expressed as remaining area during artificial alteration by “air” and “biogas” method At the initial time of the corrosion with both alteration methods, the oils showed an acidification (AN) of 2 to 2.5 mg KOH/ g and a base reserve (TBN) of about 6 to 7 mg KOH/ g. These results give evidence that the sensor is sensitive to a certain oil condition reached after a defined stress level independent from the time needed to achieve this condition. In other words, the sensor is capable of condition monitoring of engine oils run with different fuel qualities. An additional correlation with data from laboratory analysis is done in chapter 3.4. 3.3. Results from field tests The usefulness of the proposed sensor concept is also illustrated by results determined by online condition monitoring in stationary gas engines. Figure 5 shows the trend of the sensor signal of a sensor system implemented into a gas engine driven with wood gas during two oil life cycles. It is noticed that the trends depicted include only periods with running engine. Nonoperating periods are not considered here. 0 20 40 60 80 100 0 500 1000 1500 2000 2500 3000 3500 Figure 5: Trend of the sensor signal expressed as remaining area during the field tests At the first cycle after about 1300 OH, a significant decrease of the sensor signal due to corrosion of the sacrificial layer was observed. The entire metal film was corroded after about 2200 OH. Before the onset of corrosion after 1300 OH, the sensor showed an almost constant signal, i.e. base signal, as no significant corrosive attack to the sacrificial layer occurred. At the initial time of corrosion, the oil was characterised by an acidification of about 2 mg KOH/ g and a TBN of about 7 mg KOH/ g. This critical oil condition was also detected during artificial alteration (see 3.2.). At the second oil cycle, the initial time of corrosion appeared clearly earlier. Here, the corrosion process started after around 400 OH which is mainly attributed to the fact that another gas engine oil was used in this oil life cycle. According to the results, it is concluded that this oil has a lower quality than that used in the first oil life cycle. Figure 4: Trend of the sensor signal expressed as remaining area during artificial alteration by “air” and “biogas” method alteration for TBN and then a few hours later for AN, respectively. In the case of “air”, the limits were not reached till the end of the experiments. Figure 3: Trend of TBN and AN during artificial alteration with “biogas” and “air” method By looking at the sensor signals recorded during artificial alteration of the gas engine oil, an obvious shorter initial time of corrosion of about 20 hours with “biogas” can be seen in comparison to “air” causing an initial time of corrosion of about 70 hours (Figure 4). Until the initial time of corrosion, the recorded signal is almost stable showing that the sacrificial layer is not substantially corroded so far. After the initial time of corrosion, the signal for the remaining area decreases rapidly due to corrosive attack of the sacrificial layer of the sensor until the entire metal film is corroded. 0 20 40 60 80 100 Figure 4: Trend of the sensor signal expressed as remaining area during artificial alteration by “air” and “biogas” method At the initial time of the corrosion with both alteration methods, the oils showed an acidification (AN) of 2 to 2.5 mg KOH/ g and a base reserve (TBN) of about 6 to 7 mg KOH/ g. These results give evidence that the sensor is sensitive to a certain oil condition reached after a defined stress level independent from the time needed to achieve this condition. In other words, the sensor is capable of condition monitoring of engine oils run with different fuel qualities. An additional correlation with data from laboratory analysis is done in chapter 3.4. 3.3. Results from field tests The usefulness of the proposed sensor concept is also illustrated by results determined by online condition monitoring in stationary gas engines. Figure 5 shows the trend of the sensor signal of a sensor system implemented into a gas engine driven with wood gas during two oil life cycles. It is noticed that the trends depicted include only periods with running engine. Nonoperating periods are not considered here. 0 20 40 60 80 100 0 500 1000 1500 2000 2500 3000 3500 operating hours [OH] sensor signal remaining area [%] field cycle 1 field cycle 2 Figure 5: Trend of the sensor signal expressed as remaining area during the field tests At the first cycle after about 1300 OH, a significant decrease of the sensor signal due to corrosion of the sacrificial layer was observed. The entire metal film was corroded after about 2200 OH. Before the onset of corrosion after 1300 OH, the sensor showed an almost constant signal, i.e. base signal, as no significant corrosive attack to the sacrificial layer occurred. At the initial time of corrosion, the oil was characterised by an acidification of about 2 mg KOH/ g and a TBN of about 7 mg KOH/ g. This critical oil condition was also detected during artificial alteration (see 3.2.). At the second oil cycle, the initial time of corrosion appeared clearly earlier. Here, the corrosion process started after around 400 OH which is mainly attributed to the fact that another gas engine oil was used in this oil life cycle. According to the results, it is concluded that this oil has a lower quality than that used in the first oil life cycle. Figure 5: Trend of the sensor signal expressed as remaining area during the field tests T+S_3_16 05.04.16 09: 01 Seite 35 36 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 AN and TBN values, only the changes of these oil properties are compared. It can be seen in Figure 6 that the critical amount of acid is equal to an increase of AN of about 0.5 to 1.0 mg KOH/ g for both gas engine oils under all conditions applied, either laboratory artificial alteration or field tests. The correlation of the sensor signal with the consumption of the base reserve (Figure 7) showed a similar behaviour as identified for acidification: a critical consumption of about 2 mg KOH/ g was observed for the gas engine oils selected. It has to be pointed out that a dependence of the critical values for acidification and base reserve on the composition of the engine oil, i. e. base oil and additives, may occur. That is why the initial time of corrosion of the sacrificial layer could vary. Consequently, it is important to know the critical initial time for a proper interpretation of the sensor signal enabling a proper decision on the optimal time for an oil change. The method of artificial alteration described has proven to be a valuable tool for a rapid but close-to-reality evaluation of the initial time as the same critical values for AN and TBN for both artificial alteration methods and the field operation were observed. The knowledge of the initial time of corrosion can be used as pre-warning for a critical amount of oil acidification reached. The end point of the corrosion of the metal film can be equalised with the time of the oil change when considering following aspects: proper adaptation of the thickness of the sacrificial layer, operating temperature and knowledge about engine behaviour. 4 Summary and conclusions A corrosion sensor measuring the corrosiveness of gas engine oils was proposed as acidification or corrosiveness, respectively, is a crucial parameter in the operation of gas engines driven with biofuels. Therefore, a sacrificial layer was exposed to the lubricant and the material loss was monitored. Static laboratory experiments with defined model oils showed that the sacrificial layer is sensitive to the active acidic compounds in the oil. Accordingly, it is essential whether the additives can neutralise the acidic compound or not. The sensor was evaluated by artificial alteration using two methods: one simulated mild conditions (“air”) in a gas engine, e. g. driven with natural gas, and the second harsh conditions caused by a non-purified or insufficiently purified biogas (“biogas”). The results from the sensor signals clearly showed higher corrosiveness of Aus Wissenschaft und Forschung 3.4. Sensor signals compared to laboratory analyses The results from the experiments under static conditions (see 3.1.) clearly showed that the metal film of the sensor is sensitive to the amount of acidic components in the oil. If there are free reactive molecules which cannot be neutralised by the oil, the sacrificial layer is corroded according to the rule of thumb: the higher the amount of acid the faster the corrosion process. The results obtained from the artificial alteration and field operations suggest that the corrosion process starts at a critical amount of acid in the oil. For the correlation with data from laboratory analysis, the sensor signals at the time of sampling were extracted and compared with the conventional oil parameters AN and TBN determined in the laboratory. As the different types of gas engine oils used showed different initial AN and TBN values, only the changes of these oil properties are compared. It can be seen in Figure 6 that the critical amount of acid is equal to an increase of the AN of about 0.5 to 1.0 mg KOH/ g for both gas engine oils and under all conditions applied, either laboratory artificial alteration or field tests. 0 20 40 60 80 100 0 1 2 3 4 Δ AN [mg KOH/ g] sensor signal remaining area [%] field cycle 1 field cycle 2 AA-biogas AA-air Figure 6: Correlation of the sensor signal with change in AN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) The correlation of the sensor signal with the consumption of the base reserve (Figure 7) showed a similar behaviour as identified for acidification: a critical consumption of about 2 mg KOH/ g was observed for the gas engine oils selected. 0 20 40 60 80 100 Figure 7: Correlation of the sensor signal with change in TBN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) It has to be pointed out that a dependence of the critical values for acidification and base reserve on the composition of the engine oil, i.e. base oil and additives, may occur. That is why the initial time of corrosion of the sacrificial layer could vary. Consequently, it is important to know the critical initial time for a proper interpretation of the sensor signal enabling a proper decision on the optimal time for an oil change. The method of artificial alteration described has proven to be a valuable tool for a rapid but close-to-reality evaluation of the initial time as the same critical values for AN and TBN for both artificial alteration methods and the field operation was observed. The knowledge of the initial time of corrosion can be used as pre-warning for a critical amount of oil acidification reached. The end point of the corrosion of the metal film can be equalized with the time of the oil change when considering following aspects: proper adaptation of the thickness of the sacrificial layer, operating temperature and knowledge about engine behaviour. 4. Summary and conclusions A corrosion sensor measuring the corrosiveness of gas engine oils was proposed as acidification or corrosiveness, respectively, is a crucial parameter in the operation of gas engines driven with biofuels. Therefore, a sacrificial layer was exposed to the lubricant and the material loss was monitored. Static laboratory experiments with defined model oils showed that the sacrificial layer is sensitive to the active acidic compounds in the oil. Accordingly, it is essential whether the additives can neutralise the acidic compound or not. Figure 6: Correlation of the sensor signal with change in AN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) 3.4. Sensor signals compared to laboratory analyses The results from the experiments under static conditions (see 3.1.) clearly showed that the metal film of the sensor is sensitive to the amount of acidic components in the oil. If there are free reactive molecules which cannot be neutralised by the oil, the sacrificial layer is corroded according to the rule of thumb: the higher the amount of acid the faster the corrosion process. The results obtained from the artificial alteration and field operations suggest that the corrosion process starts at a critical amount of acid in the oil. For the correlation with data from laboratory analysis, the sensor signals at the time of sampling were extracted and compared with the conventional oil parameters AN and TBN determined in the laboratory. As the different types of gas engine oils used showed different initial AN and TBN values, only the changes of these oil properties are compared. It can be seen in Figure 6 that the critical amount of acid is equal to an increase of the AN of about 0.5 to 1.0 mg KOH/ g for both gas engine oils and under all conditions applied, either laboratory artificial alteration or field tests. 0 20 40 60 80 100 0 1 2 3 4 Figure 6: Correlation of the sensor signal with change in AN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) The correlation of the sensor signal with the consumption of the base reserve (Figure 7) showed a similar behaviour as identified for acidification: a critical consumption of about 2 mg KOH/ g was observed for the gas engine oils selected. 0 20 40 60 80 100 0 1 2 3 4 5 6 Δ TBN [mg KOH/ g] sensor signal remaining area [%] field cycle 1 field cycle 2 AA-biogas AA-air Figure 7: Correlation of the sensor signal with change in TBN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) It has to be pointed out that a dependence of the critical values for acidification and base reserve on the composition of the engine oil, i.e. base oil and additives, may occur. That is why the initial time of corrosion of the sacrificial layer could vary. Consequently, it is important to know the critical initial time for a proper interpretation of the sensor signal enabling a proper decision on the optimal time for an oil change. The method of artificial alteration described has proven to be a valuable tool for a rapid but close-to-reality evaluation of the initial time as the same critical values for AN and TBN for both artificial alteration methods and the field operation was observed. The knowledge of the initial time of corrosion can be used as pre-warning for a critical amount of oil acidification reached. The end point of the corrosion of the metal film can be equalized with the time of the oil change when considering following aspects: proper adaptation of the thickness of the sacrificial layer, operating temperature and knowledge about engine behaviour. 4. Summary and conclusions A corrosion sensor measuring the corrosiveness of gas engine oils was proposed as acidification or corrosiveness, respectively, is a crucial parameter in the operation of gas engines driven with biofuels. Therefore, a sacrificial layer was exposed to the lubricant and the material loss was monitored. Static laboratory experiments with defined model oils showed that the sacrificial layer is sensitive to the active acidic compounds in the oil. Accordingly, it is essential whether the additives can neutralise the acidic compound or not. Figure 7: Correlation of the sensor signal with change in TBN of artificially altered oils (AA-air and AA-biogas) as well as field samples (field cycle 1 and field cycle 2) T+S_3_16 05.04.16 09: 01 Seite 36 Tribologie + Schmierungstechnik 63. Jahrgang 3/ 2016 [3] Abatzoglou, N.; Boivin, S.: “A review of biogas purification processes”, Biofuels, Bioproducts and Biorefining, Volume 3, Issue 1, pp 42-71, 2009 [4] Richard, K.; McTavish, S.: “Impact of biodiesel on lubricant corrosion performance”, SAE International Journal of Fuels and Lubricants, Volume 2, Issue 2, pp 66-71, 2010 [5] Agoston, A.; Doerr, N.; Jakoby, B.: “Corrosion sensors for engine oils - laboratory evaluation and field tests”, Sensors and Actuators B: Chemical, Volume 127, Issue 1, pp 15-21, 2007 [6] Schneidhofer, C.; Dörr N.: “Online-Ölzustands-überwachung mit chemischen Korrosionssensoren (Online oil condition monitoring using chemical corrosion sensors)”, E&I. Elektrotechnik und Informationstechnik, Volume 126, Issue 1-2, pp 31-36, 2009 [7] Schneidhofer, C.; Sen, S.; Dörr, N.: “Determination of the Impact of Biogas on the Engine Oil Condition Using a Sensor Based on Corrosiveness”, Biofuel Production-Recent Developments and Prospects, Dr. Marco Aurelio Dos Santos Bernardes (Ed.), InTech, ISBN: 978-953-307-478- 8, DOI: 10.5772/ 20288. Available from: http: / / www.intechopen.com/ books/ biofuel-production-recent-developments-and-prospects/ determination-of-the-impact-of-biogas-on-the-engine-oil-condition-using-a-sensor-based-oncorrosiven [8] ASTM Standard D4871 (2006) Standard guide for universal oxidation/ thermal stability test apparatus, ASTM International, West Conshohocken, PA, USA, DOI: 10.1520/ D4871-06, www.astm.org [9] DIN 51352 (1985) Testing of lubricants; determination of ageing characteristics of lubricating oils; increase of the Conradson carbon residue after ageing by passing air through the lubricating oil, Deutsches Institut für Normung, Berlin, Deutschland [10] DIN ISO 3771 (1985) Petroleum products; total base number; perchloric acid potentiometric titration method, Deutsches Institut für Normung, Berlin, Deutschland [11] DIN 51558-1 (1979) Determination of the Neutralization Number, colour-indicator titration, Deutsches Institut für Normung, Berlin, Deutschland [12] Biomass power plant Güssing, available from: http: / / www. eee-info.net/ cms/ EN/ [13] OelCheck - Limitwerte - Gasmotorenöle, March 5, 2015, available from: http: / / www.oelcheck.de/ wissen-von-az/ uebersichten-und-tabellen/ limitwerte/ limitwerte-fuergasmotorenoele.html 37 Aus Wissenschaft und Forschung the oil deteriorated with “biogas” in comparison to the artificial alteration “air”. In field operation, the sensor demonstrated the usefulness and applicability of the corrosion sensor. Both artificial alterations and field tests revealed that the sacrificial layer started to corrode at the same critical amount of acidification. Laboratory analyses confirmed that the initial time of corrosion happens at an increase of AN of about 0.5 to 1 mg KOH/ g in comparison to the fresh oil. This can be also correlated to a consumption of the base reserve (TBN) of about 2 mg KOH/ g. The findings demonstrate that the proposed sensor for oil corrosiveness has a sufficient and reproducible sensitivity useful for online oil condition monitoring in engines as pre-warning system. The choice of the sacrificial material and thickness of the metal layer mainly depends on the specific application determined by fuel quality and oil type. 5 Acknowledgement This work was funded by the Austrian COMET Program (Project K2 XTribology, Grant No. 849109), the European Union’s Seventh Framework Programme (FP7/ 2007-2013, Grant Agreement No. 606080, “CondiMon”, http: / / condimon.eu/ ), and was carried out at the Excellence Centre of Tribology (AC2T research GmbH). Parts of this publication were firstly published in reference [7]. References [1] EU tracking roadmap 2015, June 15, 2015, available from: http: / / www.keepontrack.eu/ contents/ publicationseutrackingroadmap/ eu_roadmap_2015.pdf [2] Luther, R.: “Alternative fuels from the engine lubrication point of view”, MTZ worldwide Edition, 2008-03 Falls Sie eine Veröffentlichung wünschen, bitten wir Sie, uns die Daten auf einer CD, zur Sicherheit aber auch als Ausdruck, zur Verfügung zu stellen. Schön ist es ferner, wenn die Bilder durchnummeriert und bereits an der richtigen Stelle platziert sowie mit den zugehörigen Bildunterschriften versehen sind. Da wir auf die Einheit von Text und Bild großen Wert legen, bitten wir, im Text an geeigneter Stelle einen sogenannten (fetten) Bildhinweis zu bringen. Das Gleiche gilt für Tabellen. Auch sollten die Tabellen unsere Art des Tabellenkopfes haben. Die Artikel dieses Heftes zeigen Ihnen, wie wir uns den Aufbau Ihres Artikels vorstellen. Vielen Dank. Bitte lesen Sie dazu auch unsere ausführlichen „Hinweise für Autoren“ (Seite 74). Aktuelle Informationen über die Fachbücher zum Thema „Tribologie“ und über das Gesamtprogramm des expert verlags finden Sie im Internet unter www.expertverlag.de Ihre Mitarbeit in Tribologie und Schmierungstechnik ist uns sehr willkommen! T+S_3_16 05.04.16 09: 01 Seite 37