Inoculum-to-substrate ratio (I/S) in the BMP test: how to choose the sample weight

I/S (ISR, inoculum-to-substrate ratio) is the ratio of the VS (volatile solids) mass of the inoculum to the VS mass of the substrate in the bottle. The guidelines recommend an I/S of at least 2 (VS/VS), and higher values for very easily degradable substrates, in order to avoid acidification and inhibition, which underestimate the result. Too high an I/S does not harm the process, but it increases the share of the endogenous background, and hence the uncertainty of subtracting the blank.

Formula

I/S = (m_inoc · VS_inoc) / (m_substr · VS_substr), where VS is expressed as a fraction of fresh mass (kg VS/kg FM), and m is the fresh mass in the bottle.

Substrate sample weight at a given I/S:

m_substr = (m_inoc · VS_inoc) / (I/S · VS_substr).

Example

Assumptions: the bottle contains 200 g of inoculum with VS = 2.5% (0.025 kg VS/kg); substrate: silage with TS 33% and VS 31% of fresh mass (0.31 kg VS/kg); target I/S = 3.

• VS mass of the inoculum = 200 × 0.025 = 5.0 g,
• VS mass of the substrate = 5.0 / 3 = 1.67 g,
• fresh-mass sample weight = 1.67 / 0.31 ≈ 5.4 g.

If someone confuses VS with fresh mass and weighs out 1.67 g of silage, the I/S will be about 10 — the test will work, but the substrate signal will be small relative to the background.

Why not too low

At a low I/S rapid acidogenesis produces acids faster than the methanogens can consume them. The pH falls, methanogenesis dies off, and methane production is underestimated or delayed (a long lag). This shows up as a large discrepancy between replicates and an incorrect curve shape. It applies especially to substrates with a high content of sugars, starch, fats and food waste.

Why not too high

At a very high I/S the signal from the substrate is small compared with the endogenous production of the inoculum. The relative error of subtracting the background grows. In addition the mass of substrate is so small that the heterogeneity of the sample affects the result. A compromise is an I/S of about 2–4 for most materials.

Practical recommendations by type of substrate

Volume and pH limits

The bottle must hold the inoculum, substrate, water and free space for gas. Too large a volume of liquid reduces the headspace and makes measurement harder. If the substrate has a low pH or high VFA, the pH is corrected beforehand and a buffer (e.g. bicarbonate) is added, noting this in the report.

Verification after the test

After the end of the test, check the pH in the bottles (it should remain in the neutral range), the VFA (low) and the absence of an unusually long lag phase. If the pH has fallen below about 6.8, the result is suspect and the test is repeated with a higher I/S.

Checklist

• [ ] VS of the inoculum measured in the same batch.
• [ ] VS of the substrate measured, taking volatile losses into account if significant.
• [ ] I/S calculated and recorded.
• [ ] The sample weight is within the limits of pipetting/weighing.
• [ ] Initial pH in the neutral range.
• [ ] A repeat test planned if acidification occurs.

The I/S values recommended in the guidelines follow from inter-laboratory experience, but it is always worth verifying them for specific substrates, especially unusual ones.

Dilution series for uncertain materials

For a new, unusual or potentially inhibiting substrate it is worth running the test at several I/S levels (e.g. 1, 2, 4). If BMP rises with I/S and then stabilises, the lowest I/S caused inhibition. If the result is constant, the substrate is safe across the whole range. Such an approach costs more bottles, but it gives a firm justification for the chosen sample weight.

Conversion from I/S to TS mass

In some publications I/S is given on a TS or COD basis. When comparing reports, check the basis, because the same sample weight gives different numbers depending on the definition. The conversion requires knowledge of the VS/TS of both components.

The most common mistakes in choosing I/S

1. Calculating I/S on fresh mass instead of VS.
2. Using the VS of an inoculum from a different batch than the one that went into the bottles.
3. A fixed sample weight for all substrates "because that is what we always do".
4. No check of the pH after the test, despite a suspicious curve shape.
5. Inaccurate weighing of very small sample weights of high heterogeneity (e.g. straw, waste) without prior size reduction.

Avoiding these mistakes is simpler than explaining low or scattered results later.