Waste Vegetable Oil Properties with Usage and Its Impact on Artisan Soap Making

Transcription

1 Brigham Young University BYU ScholarsArchive Undergraduate Honors Theses Waste Vegetable Oil Properties with Usage and Its Impact on Artisan Soap Making Jenalyn Thorpe Follow this and additional works at: Part of the Chemical Engineering Commons BYU ScholarsArchive Citation Thorpe, Jenalyn, "Waste Vegetable Oil Properties with Usage and Its Impact on Artisan Soap Making" (2018). Undergraduate Honors Theses This Honors Thesis is brought to you for free and open access by BYU ScholarsArchive. It has been accepted for inclusion in Undergraduate Honors Theses by an authorized administrator of BYU ScholarsArchive. For more information, please contact

2 Honors Thesis TITLE WASTE VEGETABLE OIL PROPERTIES WITH USAGE AND ITS IMPACT ON ARTISAN SOAP MAKING by Jenalyn Thorpe Submitted to Brigham Young University in partial fulfillment of graduation requirements for University Honors Chemical Engineering Department Brigham Young University June 2018 Advisor: Randy Lewis Honors Coordinator: Dean Wheeler

3 ii

4 ABSTRACT WASTE VEGETABLE OIL PROPERTIES WITH USAGE AND ITS IMPACT ON ARTISAN SOAP-MAKING Jenalyn Thorpe Chemical Engineering Department Bachelor of Science This thesis examines the impact of vegetable oil usage in industrial and home settings on the properties of the vegetable oil and how its usage in soap affects the final product. Waste vegetable oil (WVO) is often used to make soap as a way to be more environmentally-friendly and create soap at a low cost in developing countries. Two settings are examined: home usage (i.e. short-term, small-scale usage) and industrial (i.e. long-term, almost continuous usage). This thesis found that lightly used oil (household use) had little to no impact on the quality of the oil, except for its scent. This resulted in a bar of soap that had very similar properties to soap made from fresh oil, except for it was somewhat brittle. However, when oil was used extensively the properties of the oil changed dramatically, including a deepening of the color, free fatty acid (FFA) content increase to approximately 5%, and an acquired scent. This resulted in a bar of soap that was darker in color, and had less lather than fresh oil soap (by 1.4 on a scale of 5). The industrial WVO had the advantage of reacting quickly with the lye to make soap, allowing the soap making process to take less time, and for the bar of soap to reach a usable stage quicker, most likely because of the high FFA%. These results demonstrate that using lightly or heavily used vegetable oil in soap results in a high-quality product for a very low cost. iii

5 iv

6 ACKNOWLEDGEMENTS I would like to acknowledge the help and contribution of my team members through Global Engineering Outreach: Abbey Wilson, Connor Weeks, and Joshua Frei. They were instrumental to making (so many) batches of soap, researching, communicating with Porcón and ultimately implementing the process developed in the Granja Porcón community. I m also grateful to Randy Lewis for his mentorship; to Matthew Memmott, for agreeing to be my reader and providing support; to Laura Kneib (owner of F.R.O.G. Soap) for her willingness to share her expertise on working with waste vegetable oil (WVO) in soap-making; to Sam Thorpe, for his continued support and encouragement. v

7 vi

8 TABLE OF CONTENTS TITLE... i ABSTRACT... iii ACKNOWLEDGEMENTS... v TABLE OF CONTENTS... vii TABLE OF TABLES... ix TABLE OF FIGURES... ix 1. INTRODUCTION EXPERIMENTAL RESULTS AND DISCUSSION CONCLUSIONS REFERENCES APPENDIX vii

9 viii

10 TABLE OF TABLES Table 1: Scale for Measure of Soap Quality... 7 Table 2: Linear Regression Parameters for FFA Content vs Time Table 3: Characteristics of Industrial WVO vs New Oil Soap Table 4: Characteristics of Household WVO vs New Oil Soap Table 5: Summary of NaOH (g) to Titrate 1 ml of Each Oil Sample and Corresponding FFA% Table 6: Summary of Average HSV Values for Industrial and Household Use Oil Samples Table 7: Summary of Scent Test Perception Numbers for Industrial and Household Use Oil Samples TABLE OF FIGURES Figure 1: Saponification reaction. Image is modified from Đokić, et al Figure 2: Soap mixture at trace Figure 3: Free fatty acid content of industrial peanut oil with time, with 95% confidence interval bands Figure 4: FFA content (%) plotted against time (i.e. sample number) for household use samples Figure 5: Industrial oil samples, from clean (0) to WVO (W) Figure 6: HSV values of industrial oil with time Figure 7: A comparison of home-usage oil samples, from least used (left) to most used (right) Figure 8: HSV values over time for household use oil samples ix

11 Figure 9: Scent of industrial oil with time Figure 10: Impact of fry time on scent of household use samples Figure 11: A visual comparison of industrial WVO (left) vs. clean oil soaps (right) made with peanut oil Figure 12: A comparison of household WVO (left) vs clean oil soaps (right) made with soybean oil x

12 1. INTRODUCTION 1.1 WVO Problem Worldwide Waste vegetable oils (WVOs) are commonly produced in high quantities in many countries, with the highest being in the US (10 million tonnes per year) [1]. They are commonly produced as by-products of community living where foods are fried. In the United States, the EPA (Environmental Protection Agency) warns that vegetable oils can cause significant effects to the environment, including: fouling shorelines; coating animals and plants and suffocating them through depletion of oxygen; destroying food supplies, breeding animals, and habitats; and being toxic [2]. Large generators of used oil are required to properly store and dispose of WVOs. WVO oil consumption is also a sizable problem in Peru. Peru produced 270 tonnes of vegetable oil in 2014, with another 38 tonnes imported; an estimated third of this oil is discarded as WVO. While estimates are not precise, much of the household WVO goes down the drains or is disposed of in the trash, while much of the industrial WVO (i.e. from restaurants) goes to Bioils, a company that collects WVO across South America and sends it to Europe, where Royal Dutch Shell converts it into biodiesel [3]. Small, remote towns like Porcón, Peru (population 2000, approximately 30 km from the nearest sizeable city) do not have access to WVO collecting facilities and create such small quantities of WVO that biodiesel production is not feasible [4]. However, this WVO can be recycled to lessen environmental impacts and enhance the community by using the WVO as the primary feedstock in a soap-making process. Such a process can also provide a means for developing communities to improve personal hygiene [5]. 1

13 1.2 Saponification Process Soap is created through saponification through alkaline hydrolysis of triacyclglycerols, the primary component of vegetable oils. The alkaline reactant used in this process is commonly industrially purified NaOH or KOH, although lye can also be created using ash with somewhat inconsistent results. Triacyclglycerols are main constituents of vegetable oil, although their precise make-up of triacyclglycerols and free fatty acids varies depending primarily on type of oil and its usage. [6]. While WVO is variable in its composition, it can be used as the primary feedstock in soap-making after treatment (generally filtration to remove entrained particles) to ensure its quality. This reaction is shown below in Figure 1, where the primary reactant is the triglyceride an ester derived from fatty acids (shown by the R groups below). This is reacted by the alkali (or base or lye, here NaOH) to form glycerol (left in as part of the soap product) and soap molecules. Free fatty acids (in the form of RCOOH) will also react with NaOH to form soap molecules. The glycerol, a side product in the soap making process, is left inside the soap structure. Figure 1: Saponification reaction. Image is modified from Đokić, et al. 2

14 1.3 Soap Making in Remote Areas Soap making can be advantageous for small towns where WVO production may be variable and limited and labor is scarce, making biodiesel production inopportune. Small-scale soap making processes have been implemented in developing communities across the world [7]. Although sources for sodium hydroxide may vary in remote areas, small quantities of sodium hydroxide can be found commercially, such as through Mercado Libre (equivalent to the U.S. Amazon) in Peru. Adding a soap-making process to the community of Porcón, Peru will be advantageous in more than one way. Hand-made artisan soaps will add to the diverse handcrafted souvenirs offered to tourists visiting Porcón and offer an additional stream of revenue. Additionally, a soap-making process can help the people of Porcón to be more environmentally aware and to reduce waste (household and restaurant WVO) by using it to create valuable products. 1.4 Understanding Properties of WVO through Usage The characteristics of soap vary greatly depending on the types of oils or fats that are used to create it. Different vegetable oils have different compositions of fatty acids which affect properties such has hardness or lather. Vegetable oil also changes properties with usage. For example, fatty acids can be freed from triglyceride structures causing the smoke point to decrease; scents of foods fried can be left behind in the waste oil; the color of the oil can darken; and particulates and water molecules can be left behind from food items and become incorporated into the oil [8]. These side effects have the potential 3

15 to impact the quality of the bar of soap created and have implications for other usages (such as re-using oil again for frying or using the WVO in biodiesel production). As usage situations vary, two scenarios were studied for WVO production industrial (longterm, almost continual usage of frying oil) and home usage (limited, short-term oil usage in small quantities over a cook stove). 2. EXPERIMENTAL 2.1. Oil Characterization FFA Content Refined vegetable oils contain less than 0.05% free fatty acids, or FFA [9]. However, triglycerides can break down in the presence of water, heat, or other elements to which they are exposed during use. The free fatty acid content can be measured by titrating the oil with NaOH (1% wt) and phenolphthalein, an indicator that turns pink when the ph of the solution exceeds approximately 8.0. In the presence of lye, the fatty acids will quickly react (in comparison to the triglycerides, that react relatively slowly through the saponification process). NaOH solution is added until it is in slight excess, turning the solution pink. The amount of NaOH that must be added to the solution to neutralize any FFA present is representative of the how broken down the oil molecules have become during usage scenarios. The approximate free fatty acid content (%) in the oil is determined by dividing the ml of NaOH solution by 1.3. Titrations were performed at least two times (more if results were inconsistent with each other) to ensure consistency and representativeness for each sample [10]. 4

16 Color Color is important aesthetically and because it affects the color of the final bar. Color can be measured through an online color summarizer for different qualities of the color [11]. While there are several ways to measure the color of the samples, color was measured in terms HSV, a common system for measuring the type of color. HSV stands for hue, saturation, and value. Hue describes the color; saturation is a measure of how intense the color is, and value is a measure of how much white or black (i.e. brightness) is incorporated with the color. Saturation and value are normalized to a scale of 100, where 100 for saturation is completely saturated (i.e. vibrant and not faded); and for value is completely bright (or no black blended). Hue is expressed in terms of angle in the color circle with red being 0, followed the other colors rainbow order until 360 [11]. To measure the color of the samples, a picture of each sample was taken on a white background during the same time of the day (to keep background light constant). All samples were taken in the same type of container to keep the thickness uniform between samples, as the oil appears darker or lighter depending on its thickness. The picture was then analyzed for its color content based on HSV, using an online color summarizer that analyzes the color of each pixel in the picture taken. Average values are reported below Scent Much of the scent in WVO is eliminated during the saponification process; however, odorous WVO can be unpleasant to work with and can also carry through to the 5

17 final bar of soap. Scent can be measured by perception on a scale of 1 to 5, as shown below. Scent tests were performed twice for each unfiltered sample, and if the results were more than 1 different, the scent tests were repeated until the results were consistent. These tests were performed by one individual; however, the tests were performed for the samples randomly and blindly as to remove as much uncertainty and bias as possible. The average scent value is recorded in the Results section. 1 strong and lingering scent of fried food 2 noticeable and somewhat strong scent of fried food 3 mild scent of fried food 4 strong scent of oil, or very weak scent of fried food 5 no noticeable scent, neutral scent, or very weak scent of oil Soap Quality While soap cannot be produced for every sample of oil during usage, soap quality can be assessed for clean and completely used oil. Soap was assessed for the following qualities: hardness, lather, conditioning effect, and scent. Soaps were tested after they had reached a safe ph range (8-10). Hardness is measured by dropping a hammer from a fixed height on a screw and seeing how deeply the screw penetrates the soap. Smell is judged by smelling the bar of soap after use. Lather is based off the number of bubbles formed after lathering with soap for seconds. Conditioning is how the hands feel after washing with the soap. 6

18 Table 1: Scale for Measure of Soap Quality Lather Conditioning Smell* Hardness 1 Excessive bubbles Feels like lotion was used Smells bad or smells like oil >2 turns deep 2 A lot of bubbles A little more moisturized Neutral 1-2 turns deep 3 Some bubbles Neutral <1 turn deep 4 Very few bubbles Takes some moisture 5 No bubbles Makes hands significantly drier Small divet No deformation *The smell test was intended to also judge the effectiveness of added scents, so as no scents were added, the main concern is if the smell of the used oil can be detected Materials Non-hydrogenated soybean oil was used to assess effects of household use; peanut oil was used in an industrial setting. Although ideally these oils would be the same, soybean oil was selected for household use tests because that is what is used in the restaurants in Porcón, Peru. No industrial vendors that use soybean oil were willing to provide samples. Soaps were made in silicon molds Industrial WVO Industrial samples were collected every afternoon for a month the usage period of oil. Samples were collected at approximately the same time every day to ensure that the amount of usage between each sample is the same. Samples were not collected on Saturdays (where usage is approximately 20% that of weekdays) or Sundays, nor were these days counted in the number of days. The industrial WVO was collected from a fast food restaurant that has the fryer on and in use for approximately 74 hours every week. 7

19 The oil is used to fry primarily chicken and potatoes. 20 samples were collected. These samples were each tested for color, scent, and FFA value. Final WVO was used to make soap samples Household WVO Household samples were created by using soybean oil in frying. The oil was used to fry assorted items, such as potatoes, donuts, onion rings, and battered fish (tilapia). Samples were collected between frying like items. One batch of household oil was created, with a total 12 samples. These samples were each tested for color, scent, and FFA value. Final WVO was used to make soap samples. 2.4 Soap Creation Soap was created by following a standard procedure as outlined below. 1. The desired amount of vegetable oil (unused or filtered WVO) was weighed, generally 200 g. 2. Personal protective equipment (PPE) was donned. This included closed-toed shoes, eye glasses or goggles, apron, and gloves. 3. The corresponding NaOH was weighed. This was determined using standard saponification numbers: for peanut oil and for soybean oil [12]. The saponification number means that for every gram oil, there is the corresponding part of NaOH. Thus for soybean oil, each gram of oil requires grams of NaOH to completely saponify the oil. 8

20 4. The corresponding quantity of water was measured. Water and NaOH were used in a 1:1 weight ratio. 5. The NaOH was slowly poured into the water and agitated until the NaOH was fully dissolved in the water. Note that this is an exothermic process, so the container would often become warm to the touch. 6. The lye solution was slowly poured into the oils and agitated with a stick blender. The solution was mixed until the soap reached a phase called trace. At this point, the mixture has reached a point where one can be assured that the saponification reaction is occurring. Here the batter is thick enough that when swirled on top of the soap mixture, a small amount stays on top of the mixture for at least a few seconds. The thickness of the mixture of this point can be described by the consistency of thick cake batter. This is shown to the right in Figure 2. This generally takes 5-20 minutes. Figure 2: Soap mixture at trace. 7. When the solution reached trace, the batter was poured into molds. 8. The molded soaps were let to sit for 2-5 days. 9. The soaps were then unmolded, and let to sit for an additional 3-5 weeks. 10. The soap was tested with a ph strip (finished soap should have a ph in the range of 9-10) to ensure that the soap was safe for use. 11. Once the soap was safe for use, it was analyzed for desired properties. 9

21 3. RESULTS AND DISCUSSION 3.1. FFA Content Industrial WVO Industrial samples were titrated to determine the FFA% in each sample, and then plotted against days of use to determine the correlation between FFA and oil usage. The samples showed a strong linear correlation with time, with an R 2 value of This plot is shown below in Figure 3. Note that the 95% confidence intervals are found tightly around the data points taken. This suggests that each day of use contributes equally to the breakdown of the triglyceride molecules found in the oil. Note that the exact amount of usage each day might vary, which accounts for some of the variation seen; however, the exact amount of usage the oil received each day was not available to account for that in the trends. A full summary of titration values is found in Table 5 in the Appendix. 10

22 6 5 Free Fatty Acid Content (%) Prediction Upper 95 % Conf. Prediction Band Lower 95 % Conf. Prediction Band Upper 95 % Conf. Single Pt Band Lower 95 % Conf. Single Pt Band Experimental Data Number of Days Used Figure 3: Free fatty acid content of industrial peanut oil with time, with 95% confidence interval bands. Table 2 below outlines the linear regression parameters for FFA% as a function of time. This is in the form y = mx + b, where y is the FFA% at time x (in days), and b is the initial FFA% in unused vegetable oil. Note that the initial FFA% is %. This is close to the tabulated estimates that refined vegetable oil sold for use has less than 0.05% FFA. Note that data could not be taken on the last day (the day that the oil was discarded); however, another WVO oil sample from the same industrial setting had an FFA value of 6.73%, a value that is higher than the day 30 predication for this batch of oil (approximately 5%). This implies that there is deviation between batches. This was 11

23 not explored in this thesis; however, a more rigorous understanding of the usage of vegetable oil would require additional data points from more than one batch of oil. Table 2: Linear Regression Parameters for FFA Content vs Time Parameter Value Standard Deviation 95% CI m b The increase in free fatty acid content in the oil had interesting implications for the soap making process. Although the time to reach trace (a measure that the saponification is well under way) was not thoroughly recorded for each batch of soap, industrial WVO reached this point significantly quicker than any other batches of soap, whether using clean oil (any type) or used soy (i.e. household WVO), which needed longer mixing times or the addition of heat to reach this stage. In general, using industrial WVO the soaps reached trace in less than 20 minutes, whereas the other soaps could take up to an hour or longer (if using a whisk). The industrial WVO soaps also harden more quickly, and reached a safe ph for use earlier. This phenomenon is most likely caused by the high concentration of free fatty acids found in the industrial WVO that readily react with the NaOH and catalyze the further reaction with the triglycerides. This idea is supported by the fact that in biodiesel production, additional lye is added to account for the fact that some of the WVO will react to make soap from the free fatty acids [10] Household WVO The FFA content for household oil samples is shown in Figure 4 below. Note that there is essentially no correlation between sample number (i.e. time) and FFA%, as the 12

24 R 2 value is This is likely because the oil was used for such a brief time that the triglyceride molecules did not break down (to any significant effect) as the industrial oil did. When the household WVO was used to make soap, it behaved similarly to clean soybean oil in how it reacted with NaOH, which makes sense in terms of the relatively similar chemical make-up. A full summary of titration values used to determine free fatty acidy content can be found in Table 5 in the Appendix. 0.8 Free Fatty Acid Concent (%) y = x R² = Sample Number Figure 4: FFA content (%) plotted against time (i.e. sample number) for household use samples. Total time of frying for household use oil was approximately 2.5 hours, so each point represents roughly 15 minutes of frying. 13

25 3.2. Color Industrial WVO Figure 5 below shows the influence of usage on the oil color, where unused peanut oil is on the left and WVO is on the right. There is a visible color gradient; however, many of the middle samples are relatively similar in color. Figure 5: Industrial oil samples, from clean (0) to WVO (W). Each number refers to how many days it has been used for. Each sample was also looked at to see how HSV values changed through time, and what type of rate of change occurred. This is shown in Figure 6 below, with a full summary included in Table 6 of the Appendix. The value (i.e. brightness) of the oil samples remains constant throughout, and the hue (i.e. color) of the samples remaining 14

26 relatively constant. The saturation (i.e. vibrancy of the color) increases with time, demonstrating that the samples become more saturated in the color as time goes on. This result matches visual perception the samples become darker in color. Data points were not able to be taken for the final sample of oil of that batch; however, a data point was collected from a different batch. This is shown in Figure 6 all the way to the right (darker points). Second order polynomials are included for the sample points that were taken from the same batch. Second order polynomials were used to provide the best fit, especially as the saturation of the samples increases more in earlier stages of usage than later stages. The second order polynomial forecast to day 30 and the final points from a different batch of used oil do not match up very well. This is likely because of the differences that occur between batches of oil fried. 15

27 y = x x R² = y = x x R² = Value H-WVO S-WVO V-WVO H S V Poly. (H) Poly. (S) Poly. (V) y = x x R² = Time (days) Figure 6: HSV values of industrial oil with time. Final points (all the way to the right) are WVO samples from a different batch of industrial vegetable oil. Lines between points are to guide the eye Household WVO A comparison of oil samples taken while creating household WVO is shown below in Figure 7, with the least used sample on the left, and the WVO shown on the right. The samples do increase in darkness; however, this change is subtle and does not significantly affect the color of the bar of soap (see Figure 12). This is not surprising, as the household oil was only used for approximately 2.5 hours. 16

28 Figure 7: A comparison of home-usage oil samples, from least used (left) to most used (right). These samples were analyzed to determine the precise color differences that occur over time, as shown in Figure 8 (HSV values). A full summary can be found in Table 6 of the Appendix. As can be seen, the HSV values remained relatively constant except for the saturation level. The value and hue values (constant here) are relatively similar to that of the industrial samples, confirming that the pictures were taken in relatively similar lighting and that the oils can be compared for color changes. The saturation value also starts at approximately the same value. Unlike the value and hue values, the saturation 17

29 value did change that is, the color itself did not change dramatically, but it did become more vibrant. This confirms what was visually seen the samples become darker HSV value H V S Sample Number Figure 8: HSV values over time for household use oil samples. Lines between points are to guide the eye Scent Industrial WVO The industrial sample perceived scent was plotted against time in Figure 9 below. The trend is generally linear, with the scent of the oil becoming more like the scent of fried foods with time. The overall variation seen (as given by an R 2 value of ) is most likely because of timing of taking the sample with respect to filtration (industrial vegetable oil is filtered multiple times during the day as it is used). Filtration removes particles which can hold scent in them, so samples that were taken soon after filtration most likely had less particles and therefore less noticeable scent. This is also most likely partially because of imperfect perceptions of scent. Based on the linear correlation, the 18

30 final WVO oil has a scent rating of 1.7. A full summary of scent perception values can be found in Table 7 of the Appendix. 5 Average Scent Score y = x R² = Number of Days Used Figure 9: Scent of industrial oil with time Household WVO Each household oil sample was tested to determine strength of the scent of the oil compared to what was fried in it, as shown in Figure 10 below. The strength of the oil scent varies somewhat from sample to sample this could be because of the specific items that were fried in the oil, or because of other variations. This is seen by the overall somewhat weak correlation, as the R 2 value is However, the overall trend is downward that is, the longer the oil is used, the more it smells like the food that is being cooked in it, which validates everyday experience. The scent of the oil reaches a scent almost as poor as the industrial samples by the end of its usage. A full summary of scent perception values can be found in Table 7 of the Appendix. 19

31 Average Scent Score y = x R² = Sample Number Figure 10: Impact of fry time on scent of household use samples. Significant to note, however, is that the overall scent that the household WVO reaches is close to the industrial WVO (extrapolated 2.4 scent rating compared to 1.7 for industrial WVO). This implies that household WVO retains the scent of food items more readily than industrial WVO. This could be in part because of the frequent filtrations done to the industrial oil, or the differences in usage. This could also be partially because the two scenarios use different types of oils soybean oil may more readily retain scents than peanut oil. Another possible explanation is that most of the scent is held in the particles entrained in the oil. The household use samples generally had more particles than the industrial oil samples did, as the household use samples were not filtered at all and the industrial samples were filtered some during usage of the oil. 20

32 3.4. Soap Characterization Industrial WVO Industrial WVO was used to make soaps, and compared to pure peanut oil for distinctive characteristics. A visual representation of the two bars of soap are shown below in Figure 11, with WVO soap on the left and pure peanut oil soap on the right. The WVO soap is significantly darker, with a slightly tan hue, where the pure peanut oil appears white. Figure 11: A visual comparison of industrial WVO (left) vs. clean oil soaps (right) made with peanut oil. The soaps were also compared on characteristics of hardness, scent, lather, and conditioning effect, which are enumerated in Table 3 below. WVO soap was harder, smelled approximately the same (i.e. neutral) and had approximately the same conditioning effects (i.e. dry hands slightly); however, the fresh peanut soap had significantly more lather than the WVO soap. The industrial WVO soap most likely does not retain the scent of the WVO because the scent reacts in the saponification reaction. Note that the used peanut oil bars of soap were from multiple batches of soap, some of which had additives. These were included in the data analysis as the additives 21

33 were intended only for coloring purposes, and as such were not anticipated to affect the characteristics of the bars. Although it is possible that these additives did affect the characteristics of the bar, the data trends suggest that the properties of the soap were independent of these coloring additives for all properties except perhaps lather, which has the widest range of results. All other bars of soap (fresh peanut and soybean oil soaps) were from one batch of soap, which could result in some of the characterizations of the soap being more reflective of the batch than the type of oil used to make the soap, as batches naturally vary to some degree. Table 3: Characteristics of Industrial WVO vs New Oil Soap Used peanut Fresh peanut Average* 95% CI Average* 95% CI Hardness 3.85 (13) (1) N/A Scent 1.89 (9) (5) 0.00 Lather 2.80 (10) (5) 0.56 Conditioning effect 2.00 (10) (5) 1.21 *number in parenthesis indicates the number of measurements averaged Household WVO A visual representation of the soap made from household WVO is shown in Figure 12 (left) below, with clean soybean oil on the right. The soaps appear very similar, as they have a color that is difficult to differentiate. The clean oil soap (right) appears that it may be less hard, as there are a few soap shavings on the outside that may have come off during handling. 22

34 Figure 12: A comparison of household WVO (left) vs clean oil soaps (right) made with soybean oil. Table 4 below enumerates the perceived characteristics of the household WVO soap compared to fresh soybean oil. Household WVO received a 1 for hardness, because each time when performing the test, the soap splintered into parts instead of the screw sinking slightly into the soap i.e. the soap is brittle. The fresh soy soap also received a low score for hardness (2) this is lower than peanut (used or fresh) soaps, which suggests that peanut oil naturally produces harder soaps. The fresh soy soap has a slightly more positive conditioning effect (3 is neutral effect, less is slightly drying); however, overall the quality of the soaps was relatively similar between the two conditions. Table 4: Characteristics of Household WVO vs New Oil Soap Used Soy Fresh Soy Average 95% CI Average 95% CI Hardness 1 (1) N/A 2 (1) N/A Scent 2.00 (4) (4) 0.00 Lather 3.75 (4) (4) 1.52 Conditioning effect 2.25 (4) (4) 0.88 *number in parenthesis indicates the number of measurements averaged 23

35 4. CONCLUSIONS 4.1 Summary of Results and Conclusions The usage of the oil has the potential to significantly impact its qualities. For example, oil used in industrial settings (i.e. long-term, intensive use) experienced a notable change in properties, including increase in FFA content to approximately 5%, a significant darkening of color, and the acquiring of scent of fried food. However, household WVO remained largely the same as fresh oil for all characteristics except for scent, which readily reached approximately the same scent of industrial WVO. These changes in the oil had an impact on the quality of the soap that they were used in. For example, industrial WVO soap was darker in color than soap made from fresh peanut oil, and was harder; however, fresh peanut oil soap had more lather. The qualities of scent and condition effect were not significantly affected between the two. The industrial WVO had the very positive effect of accelerating the saponification reaction, making it easy to make soap without the use of an electric mixture or the addition of heat. 4.2 Implications for Soap-Making in Porcón, Peru We anticipate that the WVO in Porcón, Peru will be similar to WVO created, as the soybean oil used in the restaurants is discarded daily. As such, the oil will likely still have the scent of the fried foods (primarily fish and potatoes), will be a slightly darker color, although will have the same chemical make-up. If this is the case, the WVO will create soap that is very similar to soap made from fresh soybean oil, although it may be brittle and have somewhat less lather. These poor qualities may be partially overcome by 24

36 additions to the soap of salt (for hardness) and sugar (for lather), as well as making soaps with a mixture of oils instead of one type. These changes could be explored more to confirm that these changes would be effective. Overall, we anticipate that soap from WVO in Porcón will give a high-quality product that will appeal to tourists (their intended use for the soap). 4.2 Recommendations for Further Experimentation Given limited resources and the time-consuming nature of the data collection, the quantity of data collected was limited. To give more confidence in the results presented in this thesis, the effects of usage on vegetable oil quality should be studied more rigorously. For home use samples, we recommend taking samples as oil is used in frying from multiple individuals (and cultures) that have assorted styles of frying and different foods that are fried. Additionally, for industrial samples, we recommend taking samples from multiple fryers in the same usage, as well as from different restaurants. A broader data set would give greater insight into the impact of usage on oil quality, as well as the types of usage that most drastically affect the oil quality. Additionally, as the oil types were different in the usage scenarios (peanut in industrial, soybean in household), comparison between usage impacts between the two scenarios is not perfectly valid. As such, we recommend sampling oil usage where the type of oil used is the same. Other studies could also be done to analyze the relative resistance of several types of oil to breakdown; however, these studies should be done when comparing similar usage scenarios. 25

37 Also, characterization of soap properties was done with only 3 individuals; further testing of the soaps from a varied audience would give greater confidence in the reported soap characteristics in the report. We also recommend making multiple batches of soap with the same type of oil to ensure that the characteristics of the soap are not batchspecific, but are oil-specific. Finally, additional work could be done to connect some of the properties discussed above. For example, correlations and relationships could be developed between the FFA content of the sample, its saturation, and the amount of mixing time for the soap to reach trace. These correlations would allow for an easy way to predict the performance of the oil in a soap-making setting. 26

38 REFERENCES [1] Đokić M, Kesić Ž, Krstić J, et al. (2012), Decrease of free fatty acid content in vegetable oil using silica supported ferric sulfate catalyst. Fuel 97: doi: /j.fuel [2] (2017) Vegetable Oils and Animal Fats. In: EPA. Accessed 27 Mar [3] (2016) El aceite de cocina tiene futuro. In: LaRepublica.pe. larepublica.pe/domingo/. Accessed 27 Mar [4] Historia y Cultura. In: Granja Porcon. Accessed 27 Mar [5] Ejemot-Nwadiaro RI, Ehiri JE, Arikpo D, et al. (2015), Hand washing for preventing diarrhoea. Cochrane Database of Systematic Reviews. doi: / cd [6] Đokić M, Kesić Ž, Krstić J, et al. (2012), Decrease of free fatty acid content in vegetable oil using silica supported ferric sulfate catalyst. Fuel 97: doi: /j.fuel [7] Hazeltine B (2001) Chapter 7 - Household Technologies. In: Bull C (ed) Field Guide to Appropriate Technology, 1st edn. Elsevier Science & Technology, pp [8] Beck L (2017) 'Smoke point' matters when cooking with oil. In: The Globe and Mail. Accessed 27 Mar [9] Wang T (2011) Soybean Oil. In: Gunstone FD (ed). Vegetable Oils in Food Technology: Composition, Properties and Uses, 2nd edn. Wiley-Blackwell, Ames, IO, pp

39 [10] Blair G Titrating Oil. In: Utah Biodiesel Supply. Accessed 27 Mar [11] Krzywinski M Image Color Summarizer. In: Image Color Summarizer - RGB and HSV Image Statistics. Accessed 3 Apr [12] Dunn KM (2010) Scientific Soapmaking: The Chemistry of the Cold Process. Clavicula Press, Farmville, VA. 28

40 APPENDIX Table 5: Summary of NaOH (g) to Titrate 1 ml of Each Oil Sample and Corresponding FFA% Industrial Samples Household Samples Sample NaOH (g) Average NaOH (g) Average FFA% Sample NaOH (g) Average NaOH (g) Average FFA% WVO

41 WVO

42 Table 6: Summary of Average HSV Values for Industrial and Household Use Oil Samples Industrial Samples Household Samples Sample H S V Sample H S V

43 Table 7: Summary of Scent Test Perception Numbers for Industrial and Household Use Oil Samples Industrial Samples Household Samples Sample Average Sample Average