Development of Thermal Insulation Material Using Coconut Fiber to Reuse Agricultural Industrial Waste

This paper aims to report a development of thermal insulation material using coconut fiber, which is one of agricultural industrial waste in Indonesia and a verification of an effect of thermal insulation. Seven test pieces were developed under various conditions (amount of fiber, two types of binder, amount of water, temperature setting and drying time) and thermal properties such as thermal conductivities were measured using microcomputer “Arduino UNO” equipped with radiation thermos-camera. As a result, the insulation material which mixed 30g of the coconut fiber, 30g of tapioca starch and 300ml of water, drying under the setting temperature of 120 C for 24 hours showed the lowest conductivity of 0.104 Wm K . Keywords— thermal insulation material; coconut fiber; agricultural industrial waste; reuse and low energy building material


I. INTRODUCTION
In recent years, economic growth in South East Asian countries is remarkable. According to World Economic Outlook Databases published by IMF, economic growth rate based on GDP in Indonesia shows averagely 5.4 % since 2000 [1]. Improvement for quality of life, especially indoor air quality causes increasing huge number of air-conditioner demands and affects energy consumption in household sector [2][3]. Sukarno et al. reported that 92 % of Padang electricity supply went for residential sector and 17 % of the household electricity consumption was used for cooling load [4]. It is important to effectively use of air-conditioner to reduce energy consumption. However, because typical residential building constructed by bricklaying is not attached thermal insulation material on walls, the construction brings an increase of cooling load. Therefore, to reduce energy consumption with the effective use of air-conditioner for residential buildings, development of thermal insulation material is important. Our final goal in this study is to develop thermal insulation material using domestic materials and techniques in Indonesia to spread widely the insulation material to buildings and reduce energy consumption. Additionally, one of the social issues in Indonesia is to reuse agricultural industrial waste such as the huge amount of coconut fibers discharged from coconut shells after extraction of palm fruit.
In this study, to propose reusing the industrial waste effectively, we developed thermal insulation material made from coconut fiber. In order to verify an effect of the insulation materials, we also developed simple and low-cost measuring system for recording temperature to calculate thermal conductivity using microcomputer board "Arduino UNO" as a sensor node with non-contact radiation thermos camera to measure multiple pointed temperature simultaneously and small PC board "Raspberry Pi3" as a data logging platform. To develop the measuring system, authors have ever experienced developing wireless electricity measuring system [5][6][7][8][9].

II. MATERIAL AND METHOD
The whole process for development of thermal insulation is explained as follows: 1) making test pieces, 2) developing measuring system (hardware and software), 3) calculating conductivity, 4) calibration for the developed system. Fig.1 shows a process of making test pieces. There are main four processes to make test pieces as followings; 1) Shredding fiber in 3 mm to 5 mm by cutting machine after flaking coconut shell, straining off water, 2) Mixing fiber with 2 types of binders such as natural binder, tapioca starch or chemical binder, polyvinyl alcohol and water, 3) Molding test piece whose size of width 100 mm x depth 100 mm x height 30 mm using wooden form, 4) Drying by electric oven and demolding test piece.   (Fig.3) for sensor controlling PC board "Raspberry Pi3" (Fig.4) for data logging from Micro-Computer board, two IC precision centigrade temperature sensors LM35 of Texas Instruments Inc. (Fig.5), and non-contact radiation thermos camera D6T-44 of OMRON Co.Ltd (Fig.6) which can measure thermal array of 16 points (4 x 4) (Fig.7).

A. Preparation of test pieces
2) Software of measuring system: Fig.8 shows data flow from sensor. One of the features of Arduino is equipped with communication ports to read several types of data interactively and simultaneously. For instance, in the system, analog data from LM35 as temperature data is an input to Arduino  To change these analog data from each sensor to temperature value, Arduino is written equations coding by Arduino IDE using Arduino language similar to C++ language. Data reading by Arduino is sent to Data platform in Raspberry Pi3 and converted to .TXT or .CSV format as output file using JAVA based programming language Processing 3. Each sampling code was referred author's website [11].

E. Calibration of measuring system
To calibrate of the measuring system, Japanese cypress was measured because Japanese cypress is common building material and measured with reliable conductivity value of the cypress published by Timber Industry handbook [7]. The size of the test piece for calibration is width 100 mm x depth 100 mm x height 30 mm (Fig.9) as same as other test pieces for thermal insulation. The test piece was heated until becoming heat balanced steady-state. The steady-state is presumed a state which average difference of temperature for 10 minutes is less than 0.1 o C based on the limit of resolution of the sensor.
The result is shown in Fig.10 and TABLE Ⅲ . The conductivity was 0.140 Wm -1 K -1 in almost 1 hour later when the heat balance become steady-state. According to the Timber Industry handbook [12], mean conductivity of Japanese cypress is 0.120 Wm -1 K -1 . An error ratio of conductivity between the measured value and referred value was 14.2%.  The value of conductivities is distributed from 0.104 Wm -1 K -1 to 0.160 Wm -1 K -1 . Figure.11 shows surface membrane made from tapioca starch on test pieces. An area of the membrane is marked a black line on Figure.11. All test pieces covered membrane hardened tapioca starch partly or the whole surface.
To reveal what factor affects change of conductivity, the best-performed result of 0.104 Wm -1 K -1 in Case 7 and the worst-performed result of 0.160 Wm -1 K -1 in Case 6 were compared. The components of coconut fiber, water, and tapioca starch used the same amount, however, drying time and temperature were different. Drying temperature and time of Case 6 were 40 o C and 72 hours in the oven respectively, on the other hand, the condition of Case 7 was 120 o C as 3 times higher than Case 6 and 24 hours or one-third of Case 6. The both cases were covered the membrane whole surface with visual observation. The surface of the test piece in Case 6 drawn by a broken line is uneven more than the surface of test piece in Case 7 because some parts of fiber have gravitated in hardening fiber process. On the other hand, the surface of the test piece in Case 7 was formed comparatively flat and smooth because gravitating the fiber could be prevented when the test piece become hardening under high temperature and brief time. As a matter of course, the form without unevenness can reduce heat loss by heat transfer through convective flow around the surface.
Additionally, the conductivity of test piece using PVA binder in Case1 was valued 0.105 Wm -1 K -1 as one of the best results in other cases. The membrane was generated on the surface of the piece as similar to Case 6. Therefore, effective use of PVA as chemical binder should be considered to improve conductivity.

IV. CONCLUSIONS
In this study, we developed thermal insulation material using coconut fiber to reuse industrial waste and to reduce energy consumption and simple thermal conductivity measuring system using microcomputer board Arduino UNO and reasonable PC board Raspberry Pi 3. As a result, the developed insulation material which mixed 30 g of the coconut fiber, 30 g of tapioca starch, 300 ml of water, under the dry condition with 120 o C and 24 hours was bestperformed 0.104 Wm -1 K -1 of thermal conductivity. Moreover, as drying condition for insulation material was  Fig.9 Test piece of Japanese cypress brief and high temperature, surface membrane from tapioca starch was formed effectively and functioned to increase conductivity. The developed measuring system could measure thermal conductivity with an error less than 15 % using noncontact thermos camera. As prospects for the future, to improve the insulating performance of coconut fiber insulation, we will examine 1) to combine antiseptic to prevent mold because tapioca starch would cause generating mold in some moisty cases, 2) to develop lower-environmental load insulation material using eco-friendly PVA binder.