Experimental

Introduction
Obtaining valid and reproducible data from chemical, physical, and particularly biochemical analysis can be quite a challenge especially when dealing with such complex and heterogeneous materials as organic waste, manure, and bioslurry. A thorough introduction to the importance of performing quality analytical work and assessing uncertainties is given in the CITAC guide no. 4 Quantifying uncertainties in analytical measurements. This level of abstraction may seem somewhat ambitious especially for someone new to analytical work, but remember that you as an experimenter has the freedom to choose the experimental setup that suits your immediate needs. You may be in a situation, where you do not have access to a certain piece of (specialised) equipment, but remember that the principles and the goals remain the same.

This page is dedicated to analyses relevant for the biogas entrepreneur, whether he/she is an experienced experimenter or new to laboratory work. The aim with the notes and guidelines presented here is to keep things simple without compromising scientific validity and data integrity but at the same time provide enough background information and external references so that anyone can dig further into the methods and their fundamental principles. It all depends on your level of ambition and personal motivation. With common sense, a steady hand, and a bit of algebra most of the problems encountered in the anaerobic digestion laboratory can be solved relatively easy, fast, and in a satisfying manner.

A good place to start is by reading some of the many accepted standard protocols and national guidelines that are used by the biogas community and referred to by most professionals. Recognising their strengths and weaknesses requires some experience, but luckily experts on the subject usually express their concerns in technical reports and scientific articles. An example of such a valuable piece of technical criticism is given in this article published by professor Angelidaki and colleagues in 2009. It specifically reviews some of the common standard methods and reflects on the quality of batch digestion assays; a common analysis, which will be introduced to you in the following. A very elaborate evaluation of errors associated with biogas batch experiments is given in this article from 2009 by Walker and colleagues (requires subscription to Journal of Bioresource Technology though).

Existing standards and comprehensive experimental protocols
The National Renewable Energy Laboratory under the United States Department of Energy has developed a set of standard biomass analytical procedures, which are made available to the public. Although intended for use in projects related to ethanol, many of the methods and principles described can be adapted and used in the context of anaerobic digestion. Specifically, the procedures for dry matter and ash determination are worth looking into, but the protocols for compositional analysis and fermentation are also quite interesting.

The bilingual (German/English) guideline VDI 4630 “Fermentation of organic materials - Characterisation of the substrate, sampling, collection of material data, fermentation tests” issued in 2006 by the German Association of Engineers is a recent – and very comprehensive – guideline for planning, conducting, and evaluating batch assays. It employs the usual high level of German precision and leaves the reader with a good feeling that this protocol is a structured and authoritative guideline. However, it raises one major issue that has not been properly addressed in previous protocols, namely the problem of obtaining representative samples.

Another recent technical guideline is this report from 2010 titled "Residual biogas potential test for digestates"published under the Waste and Resources Action Programme in the United Kingdom. It goes into very much technical detail on how to plan, perform, and evaluate experiments. However, like for the German guideline, hints on how to obtain representative samples and performing representative mass reduction have not been included.

Available data
A number of technical reports, scientific articles, and other documents have been issued over the years with data on various biogas experiments. In the project CROPGEN a relatively large database with more than 700 entries was developed. It is available here (Microsoft Access database).

Fundamentals of batch digestion experiments
A functional batch reactor setup can be assembled using simple and cheap means. This first example, however, uses standardised laboratory glassware for didactic reasons.The working principle in the setup shown in figure 1 is the following: the anaerobic digestion process takes place in a standard round-bottom flask (3). Temperature fluctuations inside the reactor are minimised by immersing the reactor in a water bath (4), which is temperature controlled (1). Equally important is mixing, continuous if possible (provided by the mixing rod (2) and the magnetic stirrer (1)), which ensures that substrate molecules and bacterial consortia come into close contact as often as possible. Moreover, crust formation is reduced and the produced biogas easier escapes the aqueous phase, when the liquid is disturbed. The rugged clamp and the supporting stand (5) ensure mechanical stability, especially if the level in the water bath is higher than the level in the reactor (buoyancy). The biogas (mixture of several gas components, mainly methane and carbon dioxide) is lead to a gas scrubber (9), which contains an alkaline solution that cleanses the gas for carbon dioxide. Note that the biogas tube is below the liquid surface. At the end of the tube there is a diffuser unit made from sintered glass. This results in the creation of small bubbles in (9) with large surface/volume ratio, which is important to ensure fast and effective mass transfer to the liquid. The clean methane gas then goes into the next flask (13), where it pushes on the liquid surface, since the methane enters the flask above the liquid. The displaced liquid enters the graduated beaker (15). It goes without saying that the system must be proven to be as gas proof as possible. Clamps (6), (10), and (12) hold flasks and stoppers together (remember to add a bit of laboratory-grade grease on the inside of the bottlenecks). Using the clamp (8) one can make sure that no air enters the reactor, if maintenance on the system is required. Clamp (11) comes in handy, when liquid from the beaker (15) is to be transferred back to the flask (13). Gas composition in the reactor headspace can be assessed by drawing periodical samples via the gas sampling port (7), which is made up by a thick rubber septum.



A few notes on the colourful solutions; they actually serve a purpose besides impressing visitors coming to your laboratory. The alkaline carbon dioxide scrubber solution (9), which is a 1 molar sodium hydroxide solution (40 g sodium hydroxide per 1 L of water), contains a small quantity of phenolphthalein, which is a pH indicator that turns pink/violet in dilute solutions having a pH above 8.2. Below a pH of 8.2 it becomes colourless. In other words, the colour of the carbon dioxide scrubber solution will fade as is becomes "saturated" with carbon dioxide; when the colour is about to disappear it is time to change the scrubber solution. The liquid in the displacement bottle (13) has a few drops of methyl orange added to it; it makes is a lot easier to read the volume in the graduated cylinder (15).

An example of a laboratory setup, where the principle sketched in figure 1 has been duplicated 6 times, is shown in figure 2. The reactors are 500 mL flasks, the washing flasks containing alkaline sodium hydroxide solution have a volume of 250 mL each, and the displacement bottles filled with coloured indicator solution have a volume of 1 L. The graded cylinders, made from plastic, have a volume of 500 mL and a resolution of 10 mL (distance between the lines). A very practical feature is that the reactor bottles can be replaced in virtually a few seconds due to the standard fittings. The system can be made operational in about an hour and requires daily attendance (volume readings and refilling of the displacement bottles).



It can be assembled using off-the-shelf components from any major supplier of laboratory equipment, but it is very pricy. The 2 kW heater shown costs around USD 1.000,-, the high performance magnetic mixing system is USD 1.500,-, the glass ware sums up to USD 700,-, and suitable biogas proof tubing adds another USD 100,- to the costs. So roughly USD 550,- per reactor - and that is without automated reading of neither gas production nor the other process parameters mentioned earlier. If you are lucky, you can probably get away with building one or two of these systems, but then the investment simply becomes too big.

So, let us simplify things a bit and make it affordable. Oftentimes it is useful to have a large array of reactors (say 24 or 48 units) working simultaneously in order to save analysis time.

Most researchers actually skip continous mixing and instead shake the bottles frequently; usually once or twice a day. Another major cost reduction potential lies in connecting several water baths, if the heater is equipped with a pump head (for instance paddle wheel). An example of such an installation using 250 mL brown bottles as reactors is shown in figure 3. The much smaller reactors produce a lot less biogas and a reading every two days is sufficient. The complete system consists of 48 bottles distributed in four water baths. It took about 16 hours to prepare (cutting glass pipes, poking holes in rubber stoppers, and connecting flexible hoses). Created in 2006 and still in operation today.



At the biogas plant in Blaabjerg in Nr. Snede, Denmark, the staff are using a homegrown system for evaluation of the biogas potential of the substrates they receive. Scrubbing is omitted and so the sum of both methane and carbon dioxide is obtained. The liquid displacement is taking place in graded plastic cylinders initially filed with water and then immersed upside-down in a fish tank filled with water. The biogas enters the cylinders via a gasproof connection screwed into the bottom of the cylinders. Biogas production takes place in regular 1 L flasks. In figure 4 the fish tank is seen to the right and the water bath is spotted in the lower left corner.



The above examples all require a lot of manual work both during installation/startup and daily reading of the gas production. It is surprisingly easy to rupture the fragile hoses and connections thereby introducing gas leaks. Another approach is to automate the gas measurements.