Requirements for Biomass Sampling
Guest article by Paul Janze, Sandwell Engineering
Introduction
With the current emphasis on the use of biomass for `green’ energy purposes, the importance of having good quality `hog fuel’, cannot be over-emphasized. And to ensure good quality fuel, good sampling procedures must be followed.
Woody biomass in chip form has been utilized by the pulp and paper industry for many decades and chip quality has long been recognized as having an important effect on pulp quality; to make good pulp you need good chips. Likewise for a biomass-fired plant to operate efficiently, it needs a reliable, constant supply of consistent quality fuel.
Good quality control relies on proper sampling, which must be accurate and precise and truly represent the main body of biomass. Without good sampling, quality control is based on false information. Bulk materials are difficult to sample properly on a production basis. Manual sampling can be done but it is labor intensive, prone to errors and does not easily fit into a production environment.
Collecting samples of biomass is one task not always done well, and unless the sample is taken properly, it will not be a true representation of the main product flow.
Biomass is not easy to sample. It appears in a myriad of species, forms and sizes; it knits together, doesn’t flow well, consolidates and packs easily; it can have a wide range of moisture contents, basic and bulk densities and calorific values; it will freeze; it is very dusty, catches fire easily and is self-combustible; it can contain all manner of contaminants; it can be quite fragile and care must be taken not to degrade the material.
To be of value, biomass samples need to be unbiased, accurate, precise and representative of the main lot or consignment of material.
The intent of this paper is to summarize the requirements that must be met in order to provide representative chip sampling and to do so in an easily understood and achievable manner. This paper does not consider chip classification or analysis.
1. Features of Good Sampling
Accuracy – is the ability to obtain samples which represent the true nature of the material supply.
Precision – is a statistical term relating to the number of samples taken from a lot or consignment.
Biomass is a highly variable product and requires a large number of samples in order to establish sampling precision. Precision can only be increased by increasing the number of spot samples.
A single spot sample generally is less likely to be representative of the main body of biomass than is a gross sample, which is a mixture of multiple spot samples.
2. Sampling Bias
Poor sampling can induce systematic errors that skew the results. Two common errors are:
- Spot samples are taken where certain properties are over-represented. eg. – at tail-gate of a truck where fines have settled to the bottom.
- The sample device is not capable of taking a representative sample. eg. – the sample device is too small and either rejects large pieces or overflows.
Properly designed, operated and maintained automatic samplers can minimize systematic sampling bias.
Fractionation
Fractionation or particle separation can occur from the way in which biomass is loaded / stacked and the way in which it is transported. eg. – fines will settle down to the bottom of a biomass truck over a long haul.
Additionally, fractionation can be produced by the sampling device itself; either by rejecting oversize pieces, failing to pick-up small particles, selectively picking up small particles, or breaking up particles by the sampling motion itself.
Stationary material is not uniform due to fractionation where material stratifies according to size. In order to get a representative gross sample from stationary material, a very large number of increment or spot samples are required.
3. Classification of Sampling Techniques
The quality of the sample depends upon the amount of human discretion involved, the sampling conditions, and the sampling location and timing.
- Generally, those samples which do not involve human discretion are more accurate.
- Accepted sampling conditions, include:
- Stopped Belt Cut (Laboratory Reference Method) – taking a full cross-section cut of material from off of a stopped conveyor belt.
- Full Stream Cut – taking a full cross-section cut of material from a falling stream.
- Partial Stream Cut – only taking a part of the cross-section of falling stream.
- Stationary Sampling – taking a sample from a stationary pile or container.
Generally, the stopped belt cut provides the best sample but is impractical to use in a production setting. Full-stream cut and partial-stream cut samples can produce results which are representative of the main body of material, depending upon the sampling equipment design. Stationary sampling produces the poorest sample.
- Sample location and timing
- Systematic – samples are taken evenly spaced in time or location.
- Random – samples are taken randomly spaced in time or location.
Systematic spacing generally provides better results. The challenge is to design a sampling system and procedure, which repetitively produces the most representative sample and is easy and practical to use.
4. Establishing Sampling Procedures and Selecting Sampling Equipment
Establishing sampling procedures and selecting sampling devices requires an understanding of the following:
Material Being Sampled
Biomass is highly variable in size, configuration and moisture content. It is relatively fragile and size is often an important factor, so care must be taken not to break the particles unnecessarily during sampling.
Number and Size of Samples
This depends upon the variability of the biomass and is usually determined by the plant technical department based on historical statistical results. Where new sources are coming on line, it can be expected that the number of samples required will be greater than for established sources, where historical data is available. Normal sample size required in the lab is 8-10 litres (~0.35 ftі).
Sample Collection Method
The best practical method should be used. The `stopped-belt, full-cut’ method is the best, but is not practical. The `falling stream, full-cut’ or `partial cut’ are the next best methods.
Sample Distribution Through Lot
Sample increments must be distributed through the whole volume of the lot, so that any one particle has an equal chance of being selected. This is particularly important where `fractionation’ has occurred due to fines segregation to one part of the lot / consignment.
Characteristics and Movement of Sampling Device
The opening of the sampling device must be large enough so as not to reject the largest possible piece and the capacity must be large enough to completely contain the sample without spillage. There should be no rejection by size of material or movement of the device through the material. There should be no contamination of the sample by the device.
Preferably, the sample device will pass through the entire cross-section of the stream so that each particle has an equal chance of being selected; or at least through a partial section that will contain all particle sizes within the stream.
Device speed through the flow is critical as the device must not block the flow of material. The sampling device shall be non-clogging, self-cleaning and shall be designed to facilitate inspection and maintenance. Ideally, it will not be complex or costly to purchase and maintain.
Location of Sampling Equipment
Equipment shall be located where:
- It can effectively take a representative sample.
- It is convenient and readily accessible for sample taking.
- It is readily accessible for maintenance and inspection.
Maintenance of Sampling Equipment
Sampling equipment must be safely and readily accessible to enable inspection, cleaning and maintenance. Worn mechanical sampling equipment can produce biased results, therefore it is imperative that equipment is inspected, repaired when necessary, and the performance measured regularly.
Preservation of Moisture Content
The sampling device / method shall neither dry the biomass out or add moisture.
Personnel
Sampling personnel shall be properly trained and qualified.
Criteria of Sampling Performance
Sampling procedures and equipment must be routinely monitored to ensure that the samples being collected are:
- Unbiased
- Accurate
- Provide the degree of precision required
- Representative of the whole lot
In addition, a constant sampling ratio should be maintained; ie. – constant volume or weight of sample as compared to the whole lot.
Sample Handling
Samples once collected must be:
- Clearly labeled and identified
- Sealed in moisture-proof containers
- Stored in a cool, dry place
Samples properly stored in this manner can be kept in storage for up to 36 hours with no appreciable moisture loss.
Gross Sample Size Reduction
Normal increment sample size as collected will be 8-10 litres (average bucket size); therefore, combined multiple increments (the gross sample) must be reduced down to the laboratory sample size in a manner that retains biomass representative of the whole lot.
Large gross samples shall be thoroughly mixed before being reduced to laboratory sample size.
Large samples can be reduced in size by mixing thoroughly and coning and quartering or by the use of an automatic splitter.
5. Design of Sampling Equipment
There are many `off-the-shelf’ product samplers; most of which were designed for materials other than biomass. However, it is the author’s experience that the best sampling equipment is custom designed to meet the specific requirements of the biomass sampling application considering the sampling requirements described in this paper, the product being handled, and the physical and operational constraints.
(Photo: Wood Chip Sampler)
Summary
To be of value, biomass samples need to be unbiased, accurate, precise and representative of the main lot or consignment of biomass. Biomass is highly variable in size, configuration and moisture content and is prone to fractionation and stratification, which complicate the sampling procedure.
Single spot / increment samples tend not to be accurate or representative of the main lot / consignment, particularly those samples taken from stationary loads or piles where fractionation has occurred.
Two common sampling errors, which bias results are:
- Spot samples are taken where certain properties are over-represented. eg. – at the tail-gate of a truck where fines have settled to the bottom.
- The sample device is not capable of taking a representative sample. ie. – the sample device is too small and either rejects large pieces or overflows.
The best, practical sampling procedures and equipment have the following features:
- Human discretion is minimized.
- Full-stream cut or partial-stream cut sampling is employed.
- Samples are taken systematically in time and /or location throughout the whole volume of the main lot / consignment.
- They do not introduce bias or contaminants into the sample.
- They are convenient to use.
- Preferably, they automatically reduce the gross sample to the laboratory sample size.
- Preferably, they have low capital and maintenance costs.
It is not easy to achieve all of these objectives, however, the primary goals of accuracy, precision and representativeness must take precedence. Convenience and cost, while important, are secondary considerations.
About our author
Paul Janze is a senior material handling specialist with more than 30 years experience in engineering, equipment design and manufacture, project management and plant maintenance, primarily in the forest products industry. He is a specialist with difficult-to-handle materials such as wood chips, hog fuel, wastewood and bark, biosolids sludge and wet pulp, poultry litter and boiler ash which all have differing and unique handling characteristics.
For more information contact:
Mr. Paul Janze
Material Handling Specialist
Sandwell Engineering Inc.
885 Dunsmuir Street, Suite 600
Vancouver, BC, V6C-1N5
Canada
Telephone: 604-684-9311
Fax: 604-688-5913
Email: pjanze@sandwell.com
Web site: http://www.sandwell.com/
Guest articles for the Ask Joe! Column are always welcome, for more information please contact Joe Marinelli directly at his email address: joe@solidshandlingtech.com.