Environmental Sampling and Analysis; A Practical Guide: Review and Summary of Section I: Environmental Sampling (Lawrence H. Keith, Ph.D. Lewis Publishers, 1991)

Environmental Sampling and Analysis; A Practical Guide: Review and Summary of Section I: Environmental Sampling (Lawrence H. Keith, Ph.D. Lewis Publishers, 1991)

      This is part 1 of a two-part review and summary. The second one will be on environmental analysis. Keith writes concisely and thoroughly, putting a lot of information in his words.        

   In the preface, Keith notes that the general public does not well understand the numbers in the determination of detection limits for chemicals.

“The ability to detect a minute trace of a contaminant does not necessarily make its removal technically or economically practical.”

In describing the purpose of the book, he makes the following statements:

“The goal of this guide is to provide a basic understanding of the principles that affect choices made in the many and complex steps involved in environmental sampling and analysis.”

“This is not a book on how to perform environmental sampling and analysis, rather it considers what aspects of sampling and analysis can successfully attain ‘data of a known quality.’”

     Data of a known quality simply means that the data produced from sampling and analysis meet one’s data quality objectives, whatever they are. This book is concerned with sampling and analysis strategies to meet those objectives.

 

Planning and Sampling Protocols

     The goal of sampling is to acquire a representative population. Usually, the representative samples should reflect the population as a whole. Nonrepresentative samples may be needed for comparison. Sampling and analysis often need to be carefully planned to meet the chosen data quality objectives (DQOs). DQOs convey confidence levels and consider the limits of variability in results, or how much error and uncertainty can be tolerated. A DOQ may be qualitative. An example given is having a strategy for dealing with contaminated samples. Solutions might be to discard those contaminated samples, to figure out the issue and resample, or to only use data above the background contamination. Quantitative DOQs are more specific and utilize quantitative and statistical analysis metrics like standard deviations, relative standard deviations, percent recovery, relative percent difference, and concentration. Quantitative DOQs also involve strategies to get the desired detection limits and representative detection levels. DQOs must also be developed within the sampling and analysis cost constraints. Site location and accessibility, the numbers, kinds, and complexity of sampling, and the frequency of sampling are three factors to be considered when evaluating cost. A Sampling plan checklist is shown below.

 

 

Basic Considerations of Planning and Good Sampling Protocols

     Samples can be used for multiple purposes and a sampling plan may have multiple DQOs with the final DQO being a compromise with the others. Ideally, data quality should not be compromised but realistically it is not always necessary.

     The first step of environmental sample planning is defining the objective(s). They are broadly categorized as exploratory (surveillance) and monitoring (assessment) goals. Exploratory sampling provides preliminary information. Monitoring usually provides information about analyte concentrations in place and time. Monitoring may or may not have a regulatory purpose.

“Exploratory sampling can help establish the chemical species of concern and the range of their concentrations and variability. This optimizes the selection of sampling equipment, analytical methodology, numbers of samples, and sampling protocol and establishes meaningful QC criteria.”

     Control sites are important for interpreting monitoring data. Selected control sites should have everything in common with the sampling site except the real or potential pollution source.

     Preliminary sampling is sometimes employed where a 10-15% section is sampled and monitored. Supplementary sampling is sometimes required and also is estimated at 10-15% of the monitoring effort. Data quality and quantity requirements should be balanced with budgetary limitations. The data users, samplers, and analyzers should all be involved in the process. It also needs to be decided whether the study will follow regional, state, and federal protocols or other Good Laboratory Practices (GLP) regulations. Strict protocols may be required.

     The level of data confidence and likely margins of error should be determined. Levels of sampling detail are often budget-constrained. Whether the sampling and analysis will be done under regional, state, or federal protocols needs to be determined. How the data will be analyzed also needs to be determined. Judgmental, random, or systematic sampling approaches may be used. There are random variable methods such as hypothesis testing, estimation interval, tolerance interval, control charts, etc. These all require random sampling. Systematic sampling is preferable for geostatistical analysis. Time and space are incorporated and interpolating algorithms like kriging may be employed as they are in mapping. Judgmental sampling involves “purposely trying to obtain the highest concentration of analyte possible.” All of these methodologies should be documented in reports. Important considerations are the types and numbers of QC samples. This is dependent on the DQOs and the nature of the study. The sampling protocol should include procedures for collection, packaging, labeling, preservation, transportation, storage, and documentation. Files preparations like filtering and pH adjustment and additional measurements like dissolved oxygen attained with a dissolved oxygen meter) may be required. The analytical method to be used can dictate the sizes of samples needed and other factors such as preservation and selection of storage containers.

 

 

Basic Sampling Approaches

     As noted, there are three primary approaches to sampling. These are stratified (judgmental) random, systematic random, and systematic judgmental. Usually, they are called judgmental, systematic, and random. Random sampling removes the most bias but also requires the most samples. The table below shows some characteristics of each.

 

 

 

Primary Sampling Approaches

 

Approach        Relative Number of Samples       Relative Bias        Basis of Selecting Sample Sites

 

Judgmental        Smallest                                                   Largest                    Prior history, visual assessment

                                                                                                                                          of technical judgment

                                                                                                                                        

Systematic         Larger                                                        Smaller                    Consistent grid or pattern

 

Random               Largest                                                      Smallest                  Simple random selection

 

     Some random sampling is usually desirable. Sample sites with larger volumes require more samples in general. Homogeneity or heterogeneity of sample sites is also important as heterogenous sites will require more samples.

 

“Often a combination of judgmental, systematic, and random sampling is the most feasible approach; however, the sampling scheme should be sufficiently flexible to permit adjustments during field activities.”

 

     Sample planning should also consider the migration of pollutants. As much knowledge as possible about the potential migration of contaminant plumes in air, water, sludges, etc. should be considered. Measuring contaminant concentrations at various distances from the source is commonly employed in sampling. Possible migration of contaminants to nearby control sites where ‘blank’ samples would normally be collected is also an important consideration.

 

     Contamination during sampling needs to be prevented.

 

“Contamination by sampling g devices and materials can contribute relatively large errors in comparison to analytical procedures, especially when the analytes of interest are at low concentrations.”

 

     Reactive or sorptive materials should be avoided. Any material that could leach potential contaminants should also be avoided. Tubing, gaskets, and metal and plastic components need to be resistant to reactions, desorption, or leaching.

     Sampling devices and protocol should favor the most labile analytes or those that change the fastest in the time between collection and analysis. Lability is dependent on reactivity, volatility, and sorption potential. Replicate samples must often be collected, prepared, and preserved differently.

 

 

 

Composite Sampling

 

     Care must be taken in planning for composite sampling, which involves combining portions of multiple samples. It is often used to reduce the cost of analyzing samples by reducing the amount of samples. Composite sampling may be an applicable strategy in the following situations:

 

1)        When samples taken from different locations or populations are analyzed to determine if the component of interest is present

2)        When aliquots of extracts from various samples composited for analysis are analyzed to determine whether the component of interest is present

3)        When representativeness of samples taken from a single site or population needs to be improved by reducing intersample variance effects {due often to heterogeneity}

4)        When representativeness of random samples, removed from a potentially heterogenous matrix, needs to be ensured by reducing the effect of variance between aliquots

5)        When a necessarily limited size of the material available for analysis, such as blood specimens, needs to be increased to achieve analytical performance goals

6)        When estimating the frequency of a trait, such as the HIV virus, to reduce the cost and/or mean square error of the estimate while, at the same time, preserving the confidentiality of the samples individuals.

 

 

 

     Since sampling is usually much cheaper than sample analysis, more samples can be taken, examined, and considered for analysis. Composite sampling can reduce margins of error and false positives. More samples can be taken than needed to test for presence, with only those samples analyzed further to reduce analysis costs. Composite sampling can also increase the total amount of sample material available if there are limitations on getting enough sample material. The limitations of composite sampling are given below:

 

1)        When composite sampling is used, interactions among analytes or organisms must be considered carefully. Take care to ensure that analytes or organisms from different samples will neither be mutually destructive nor create analytical interferences. If corrective action is not taken when such problems are suspected, composite sampling that is not representative of any of the original samples may result in a test material

2)        When the objective of the monitoring program is a preliminary evaluation or classification, compositing may dilute the analyte to a level below the detection limit, producing a false negative

3)        If sampling costs are greater than analytical costs, analyzing each sample individually may be more cost-effective

4)        When considering multiple analytes in a composite, information regarding analyte relationships in individual samples will be lost

5)        If compositing reduces the number of samples collected below the required statistical needs of the DQOs, then those objectives would be compromised

 

 

 

Safety Considerations of Sampling

     Safety considerations in sampling are aimed at reducing the samplers’ exposure to contaminants. OSHS requires special training for sampling at hazardous waste sites. Safety equipment is mostly personal protective equipment (PPE) for reducing exposure, which may include hard hats, safety glasses, safety boots, respirators, self-contained breathing air, gloves, or hazmat suits.  Sites with suspected high concentrations of contaminants such as landfill leachate and some other wastewater streams may require chemically protective clothing. Well-designed and well-documented safety protocols for sampling at vulnerable sites are important. 

 

 

Quality Assurance and Quality Control (QA/QC)

    The goal of quality assurance and improvement is to identify, measure, and control errors in every link of the planning-sampling-analysis-reporting chain and to minimize and correct individual errors and their cumulative effect. Sample collection, preservation, storage, and shipment must all be subject to QA/QC protocols. For analysis, labs have their own QA/QC protocols. Bias and Precision are the two parameters most associated with measurement quality objectives. Bias is defined as a systematic deviation (error) in data. It is usually caused by contamination of the samples. Precision is defined as random variation in data.

“One objective of any sampling quality assurance program is to provide the type and number of quality control samples necessary to control and minimize the effects of bias and precision in the sampling effort.”

     Statistical methods are used to evaluate both systematic and random errors. Systematic errors can be introduced by choice of sampling device, container, and preservation methods. For example, the sorption or reaction of a container into the sample can invalidate samples. They introduce bias. In sampling, but especially in analysis, bias from contamination is introduced by material transfers, filtrations, physical measurements, aliquot preparation, and spiking.

     Field QC samples such as blanks, spikes, and replicates provide a basis for comparison. They should be handled the same way as the environmental samples to reduce bias.  

 

 

Blank Samples

     Blanks are defined as matrices that have negligible or unmeasurable amounts of the substance of interest.

DQOs determine the need for blanks. When there is a possibility of introducing extraneous material into a sample or during analysis, a program of collecting blanks should be devised to measure that extraneous material. As noted, this extraneous material could cause bias. Blanks can be used in any part of the full sampling-analysis process and are named accordingly. They can compare introduced contaminant concentrations before and after each individual process.

     Field Blanks are simply analyte-free media similar to the sampling matrix. These give a background to measure against. They can measure contamination during the whole process. One field blank per day is usually recommended.

“Field blank water samples consist of triple distilled water that is carried to the sampling site and exposed to the air there so that any contamination from the air can be measured and accounted for [Cowgill, 1989.]”

     Trip Blanks are test samples taken from the lab to the sampling site and returned to the lab unopened. They are used exclusively to measure contamination from the sample container and preservative during transport, field handling, and storage. One trip blank per day is also recommended. According to the EPA cross-contamination from containers and preservatives occurs only with VOCs.

     Background Samples, also known as matrix blanks or control samples, are taken near the time and place where the test samples will be taken and where the analyte of interest is thought to be present at a background level. Matrix complexity determines how important matrix blanks will be.

     Equipment Blanks are samples of analyte-free media that were used to rinse sampling equipment. They document adequate cleaning of the sampling equipment. They are collected after decontamination and before resampling.

     Material Blanks are samples of material used in the construction of groundwater wells, for example, that may be introduced into groundwater samples. Grout or cement is often used in these wells. The material blanks document the decontamination of the materials.

If field blanks indicate there is no potential problem, then the other blanks may not need to be analyzed. They are often collected just in case potential contamination is suggested by field blanks.

When other extraction devices are used for indirect sample collection such as air filters, charcoal, an ion exchanger, or Tebax, blank samples involve an unused extraction device. Tenax and charcoal are sorbents so as blanks they must be kept sealed until analysis.

 

 

 

Background (Control) Samples

 

      He mentions two types of control samples: those used in determining whether or not an analytical procedure is in statistical control and those used to determine whether or not an analyte of interest is present in a studied population but not in a similar control population. Background samples can determine how much a selected sample site differs from the norm. Background concentrations are compared for many contaminants in many sites as the method is common and is a valid way to approach science. Collection and analysis should be the same for background samples and samples of interest.

     There are some recommendations for choosing local control sites: 1) they should be upwind or upstream of sampling sites; 2) If possible, background samples should be taken first to avoid introducing contamination from the site of interest; 3) travel between local control sites and sampling sites should be minimized to reduce chances to introduce contamination.

Field Spike Samples

     Field Spike samples are field samples in which a known amount of analyte is added during field collection. They are used to note any changes in that analyte concentration during collection, preservation, transport, or to assess matrix effects. They are often used where complex matrices are present. They should be prepared up to laboratory standards, if possible.

 

 

Sampling Water Matrices  

     “The ASTM Committee D-19 lists types of waters as surface waters (rivers, lakes, artificial impoundments, runoff, etc.), saline waters, estuarine waters and brines, waters resulting from atmospheric precipitation and condensation (rain, snow, fog, and dew), process waters. Potable (drinking) waters, glacial melt waters, steam, water for subsurface injections, and water discharges, including waterborne materials and water-formed deposits. Many of these water types require special sampling and handling procedures peculiar to that source.”

     An issue with water matrices is heterogeneity. This makes it harder to get truly representative samples. Organic compounds can occur in suspension or as layers in a water matrix. Oils and other organics may float on water. Others. Like halogenated compounds will sink to the bottom. The composition of surface waters may vary by season and by depth. Concentrations of contaminants and properties like pH can be affected by flow conditions of moving waters or water levels of less mobile water bodies like lakes. He notes the preferred sapling conditions for flowing waters, such as streams:

“When a single fixed intake point is used, it should be located at about 60% of the stream depth in an area of maximum turbulence, and the intake velocity should be equal to or greater than the average water velocity [Newburn].”  

The ocean, lakes, and some deep rivers may have thermal or chemical stratification that needs to be considered in sample planning. Precipitation (rain, snow, fog, dew) samples vary based on meteorological conditions. Automated samplers that open when precipitation begins are the preferred type of sampling device for sampling precipitation. Early precipitation of rain and snow usually has higher concentrations of contaminants early in the precipitation period, due to gathering dust, particulates, and some atmospheric gases (during falling) early in the precipitation event.

     Groundwater sampling has its own constraints. When the sampling is done relative to rain events, seasons, and agricultural chemical applications can be important. For groundwater monitoring wells, water alteration by well drillers and sampling must be minimized, A well is typically purged of 3 to 10 times the well volume to eliminate stagnant water. The hydraulic conductivity of the aquifer should also be determined.

“Groundwater vulnerability to contamination is affected by water depth, recharge rate, soil composition, topography (slope), as wells as other parameters such as the volatility and persistence of the analytes being determined [Dupuy, 1989]. In planning groundwater sampling strategies, knowledge of the physical and chemical characteristics of the aquifer system is necessary (but almost never known). Groundwater presents special challenges for obtaining representative samples.”

 

 

Water Sampling Devices

     Stainless steel is preferred for sampling devices. They are typically rinsed with triple distilled water to remove any inherent contamination. Medical-grade silicone rubber is preferred for things like pump gaskets to avoid contamination by peroxides that occur when conventional silicone rubber is used. Teflon is also a preferred material since it does not create potential contaminants. Food-grade PVC (to prevent phenolic compound contamination that may occur with non-food-grade PVC) is usually acceptable. The potential sorption of analytes of interest should be considered. Samples from large bodies of water are usually collected manually. Automatic samplers that can collect discrete or composite samples are typically used for streams and wastewater discharges. Glass vacuum pump samplers are commonly used for dew and fog. When sampling water for VOCs it is common to use glass vials with Teflon caps and to leave no headspace. These should also be used when organic compounds are the analytes of interest. If metallic compounds are the analyte(s) of interest then plastic, usually polypropylene plastic, is preferred, or glass with nitric acid added for stability.

     The selection of sampling equipment should also consider biological oxygen demand (BOD) and chemical oxygen demand (COD), especially vacuum samplers. These produce higher BODs and CODs than peristaltic pumps. Sampling time intervals should also be considered.

     Cement used for PVC pipe joints, grout, and other materials can contaminate groundwater samples.

“Groundwater sampling devices should be designed to avoid excessive aeration so that analyte volatilization and oxidation are minimized.”

     Electric submersible pumps, bailers, suction-lift pumps, and positive displacement bladder pumps (often the best choice for accuracy) are commonly used for sampling groundwater. Bailers are often used for purging and for sampling small-diameter wells. He notes that due to aeration concerns, field blanks should especially accompany bailer samples.

    Contamination of water samples is always an issue to mitigate. Devices, piping/tubing, and containers are the main culprits. Blanks are the main means to assess the contamination of samples. Many factors affect sorption and leaching and the determination of contamination can be complex. As the table below shows there are many different possible contaminants when sampling water wells including organic compounds and metals.

 

Potential Contaminants from Sampling Devices and Well Casings

Material                                                                                   Contaminants Prior to Steam Cleaning

Rigid PVC-threaded joints                                                Chloroform

Rigid PVC-cemented joints                                               Methyl ethyl ketone, toluene, acetone, methylene

                                                                                           chloride, benzene, organic compounds, tetra-

                                                                                           hydrofuran, ethyl acetate, cyclohexanone, vinyl

                                                                                           chloride

Flexible or rigid Teflon tubing                                          None detectable

Flexible polypropylene tubing                                          None detectable

Flexible PVC plastics tubing                                            Phthalate esters and other plasticizers

Soldered pipes                                                                  Tin and lead

Stainless steel containers                                                 Chromium, iron, nickel, and molybdenum

Glass containers                                                               Boron and silicon

 ___________________________________________________________________________________

     Sampling of public water supplies should be done before treatment or purification to assess the true state of the aquifer.

 

 

Sample Preservation

     Preservation may involve pH adjustment, protection from light, absence of headspace, chemical addition, and temperature control. The chemistry and reactivity of analytes of interest must be considered. Holding times vary by analyte, preservation method, and analytical methodology. EPA provides maximum holding times (MHTs) for many analytes. MHTs for VOCs are about 14 days but many can be longer.

“Water samples are in a chemically dynamic state, and the moment they are removed from the sample site chemical, biological, and/or physical processes that change their composition may commence [Parr, 1988].”

Some of the ways water samples change include volatilization, sorption, diffusion, precipitation, hydrolysis, oxidation, and photochemical and microbiological effects. Free chlorine in a sample can create chlorine by-products. Sodium thiosulfate is added to remove free chlorine. We added this to sample bottles before taking water well samples of chlorine-disinfected private water supplies. Low temperatures are desirable for water samples, especially when they are to be tested for microbiological contaminants. We used sampling carry bags with ‘blue ice’ to keep them cool and stored them in a cold refrigerator while waiting for the lab to collect them for analysis, usually that day or the next day.

“Analytes may also form salts that precipitate. The most common occurrence is precipitation of metal oxides and hydroxides due to metal ions reacting with oxygen. This precipitation is usually prevented by adding nitric acid.”

 

Sampling Air Matrices

     Analytes of interest in air tend to be reactive. They may also be between phases such as gases and solids or gases and liquids. Sampling often changes those phase relationships which makes most air sampling tenuous, according to the author. Air samples may be of indoor air, ambient (outdoor) air, air from stacks or emissions exhausts, or air from soils. These soil vapor samples are often used to detect and assess contamination around landfills. Air sampling typically uses solid sorbents for vapor phase compounds and filters for solid or particulate phase compounds. Analyte concentration sin air samples tend to vary considerably over short periods of time which complicates air sampling.

     Getting representative samples is a common problem with air samples.

“Vapor pressure and polarity are two of the most important physical properties affecting compound sampling from air. Important contributors for obtaining representative samples include the efficiency of the collection apparatus, integrity of the sample entering and being removed from the apparatus, location of the sampler, and timing of the collection [Ziman].”

Since stack emissions tests are samples caught at one point in time, they can be erroneous. Quality control measures include taking duplicates and taking screening samples ahead of time. Getting representative indoor air samples may require evaluating building HVAC systems, consideration of sampler location, and knowledge of chemicals used in the vicinity. Reactions with filter or container material are other important considerations. Sorbent sampling is wrought with representativeness issues including capture efficiency and recovery efficiency.

     Filters are used to sample aerosol particles. Solid sorbents are used to collect VOCs and SVOCs. Sorbents are generally of three types: organic polymers, inorganic sorbents, and carbon sorbents. Steel canisters with interior walls electropolished are commonly used to collect ambient air samples. Teflon may also be used. Soil vapor samplers utilize perforated plastic tubing inserted into bored holes. They are extracted with peristaltic pumps into Teflon bags. A second level of sorbents may be able to determine if a breakthrough of VOCs/SVOCs may have occurred, which would lower the concentration of the sample. For SVOCs, a particle filter is usually placed in front of the sorbent cartridge. Solvent extraction or thermal desorption may be used to recover organic compounds. When I worked on some coring projects, we selected parts of some shale cores or coalbed methane cores for desorption analysis in desorption canisters. In those projects, we were mainly concerned with how much and how fast natural gas was desorbing out of the rock to determine the %, rate, and ultimate recovery of desorbed gas.

     Nonvolatile compounds are usually bound to solid particles with negligible concentrations in vapor phase. Dry deposition has its own sampling requirements. Impingers (bubblers) are used in some air sampling but have drawbacks due to evaporation.

     Common air sampling problems include irreversible sorption on the walls of sample containers. Sampling aerosol particles can be challenging. The goal is to approximate isokinetic conditions. Aerosols are “colloidal suspensions of particles in a gas [Lodge].”

     Air samples may be strongly affected by meteorological conditions during sampling. Wind direction is the most important factor. Others are wind speed, temperature, atmospheric stability, atmospheric pressure, and precipitation. Wind speed may affect volatilization rates of contaminants from liquid sources. However, higher wind speeds may dilute those concentrations. Higher wind speeds can also carry more particles which can increase concentrations of heavier particulates. Higher temperature also increases volatilization. Atmospheric instability refers to vertical motions in air masses which disperse contaminants. Downwind contaminants are usually higher under stable atmospheric conditions. Higher volatilization rates occur when atmospheric pressure decreases. Atmospheric pressure influences volatilization in and around landfills as measured with soil vapor probes. Precipitation removes some contaminants from the air when it falls, reducing the concentration of them in the air. However, their concentrations increase temporarily in the early parts of a precipitation event.

     Topography can also affect air sampling. Mainly, it affects wind direction. There are daily changes of wind direction associated with some topography. Air pollution can collect in valleys. These changes occur in response to temperature changes which can lead to changes in wind direction. Cooled air can carry more contaminants.

 

 

Sampling Biological Matrices

     Some things to consider in sampling biological matrices are the homogeneity of the matrix, analyte concentration, extraction and concentrating efficiency, and the sensitivity of the method.

“Sampling biota for chemical analysis presents unique challenges because of the vast size differences between species, variations within a study population, species mobility, and tissue differentiation.”

     Previous exposures and bioaccumulation of the analyte of interest should also be considered and accounted for.

     Large variabilities are common with biological samples. This makes it harder to get representative samples. Trying to achieve homogenous samples is the goal of getting representative samples.

     Sampling fish requires catching them. Collecting botanical samples is usually much easier.

     Sample size can be problematic for biota samples. If the analyte concentration is very low, then the sample should be large enough to provide a measurable amount of analyte. Making samples fully homogenous by grinding and mixing can make the sample larger but also more composite.

     Preserving biological samples should be simplified. Freezing is one option to prevent decomposition.   

 

 

Sampling Solids, Liquids, and Sludges

     Sampling solids, sludges, soils, and some liquids are commonly not representative due to matrix heterogeneities. Certainty levels are usually lower. Background/control samples are often very important in sampling these matrices. The size and distribution of the sampling population are also very important. Test samples should be as large as possible. One strategy is to use a “tared bottle with the solvent or acid may be prepared so the sample can be added to it.” This means preservation begins at extraction. Composite samples may improve homogeneity. Spiking field samples may not be possible with these matrices. Statistical analysis should aid sampling and analysis.

     Sediment sampling usually involves coring or dredging a water body bottom. Core samplers maintain the vertical integrity of the layers, but bottom dredgers do not. Scoops and trowels are quick, but the samples collected are disturbed samples. However, they are good for compositing samples.

     Soil sampling devices are chosen based on soil characteristics and the analyte(s) of interest. Soil samples may be done to test for analytes from recent spills or suspected long-term contamination. Long-term contamination can warrant deeper soil samples. Scoops or shovels may be used for shallow samples if the analyte(s) of interest are not volatile. Sampling devices should be decontaminated between successive samples to avoid cross-contamination. This can produce QC samples known as equipment blanks. Tube samplers can also be used for shallow soil samples.  For soil samples greater than one foot deep, an augur is usually the best device choice. There are powered, unpowered augers and they can be used to drill to the required depth then use a split barrel sampler or a soil probe to get the sample to avoid cross-contamination and prevent the release of volatiles.

     Sampling of some wastes can be dangerous, even explosive. Minimizing disturbance during sampling, handling, and transport can prevent the loss of volatiles. Volatiles are target analytes when investigating landfills, underground storage tank sites, and oil & gas sites. I used to collect volatile component samples from drill cuttings in iso-tubes and iso-jars. The tubes would pull a vacuum to draw volatiles out of the cuttings. In time more and more volatiles are released from the cuttings into the air space in the jar. Occasionally the gas pressure would increase enough to bulge or burst the plastic jar.  

     Sample preservation involves sealing containers quickly, minimizing headspace, and refrigerating samples as soon as possible.

     In summary, the most frequent changes and major concerns with sampling these matrices are volatile loss, biodegradation, oxidation, and reduction. Cooling can preserve volatiles and slow biodegradation but freezing can cause volatiles to outgas.