S A Wilson and A Shuttleworth

The use of in ground barriers, particularly vent trenches, is a common method used to protect development from landfill gas migration. A new system for installing passive venting barriers is described which overcomes many of the disadvantages associated with conventional trench systems. It uses highly efficient geocomposite vent nodes, which are driven into the ground and connected to a collection/dilution duct, to allow safe venting to atmosphere. The system minimises spoil and contact by installers with contaminated soils and can be installed in restricted spaces.

There is little design guidance available for such barriers, because the performance depends on a complex relationship between several highly variable parameters. A simplified method is described which allows the relative performance of different vent node spacing and ventilation methods to be assessed. This should help to ensure that gas vented to atmosphere from the system is at acceptable concentrations.

A monitored trial of the system has been undertaken which demonstrates the barrier is effective in reducing migration of landfill gas and diluting methane and carbon dioxide concentrations at the vent outlets to less than 1% v/v.

Vertical in ground barriers are used extensively to prevent gas migration from landfill sites to below susceptible targets (usually a nearby development).

There are two common methods of forming a barrier to gas migration:

• Using very low permeability materials to resist gas flow,

• Using highly permeable materials to allow the gas to vent to the surface.

Current methods of forming a gas resistant barrier usually involve the excavation of a trench and backfilling with either an impermeable material such as bentonite, or the inclusion of a gas resistant membrane. Vent trenches are normally constructed using trenches backfilled with either gravel or geocomposite venting media to promote gas flow to the surface. An alternative method is to provide a series of discrete vent wells at regular spacings. These methods allow the gas to exhaust directly to atmosphere without any dilution in the system.

Recent European legislation suggests that the primary method of gas management from heavily gassing landfill sites should comprise enclosed flaring or energy utilisation. This prevents emissions of methane (a greenhouse gas) to atmosphere. Control contingencies to support the primary gas management system may include perimeter gas barriers as a secondary method of preventing off site migration.

In the past it has also been common to manage gas in the ground by uncontrolled venting to atmosphere. The Pollution Prevention and Control (England and Wales) Regulations (2000) implemented the Landfill Directive and apply to all new landfills and all existing ones from 2003. This requires the use of best available technology (BAT) and therefore the venting of undiluted gas to atmosphere should be avoided wherever possible.

It is, therefore, considered unacceptable to passively vent gas to atmosphere which contains greater than 1% v/v methane or 1.5% v/v carbon dioxide, primarily on health and safety grounds. One implication of this is that vent trenches must dilute gas to tolerable levels before discharge to the atmosphere.

The concept of the passive dilution barrier is to form a low pressure area relative to the surrounding gassing ground, to encourage gas to flow towards the barrier. This is achieved by driving discrete vent nodes into the ground, which are connected to a collection/dilution duct running along the top of the strips. The nodes comprise highly efficient geocomposite strips. The duct has a high flow of fresh air through it by means of passive ventilation. This is one of the key advantages of the system as it:

• dilutes gas emissions to tolerable levels,
• reduces pressure and causes a suction effect in the geocomposite vent nodes, which enhances gas flow from the ground towards the vents.

Ventilation of the duct can be achieved using a combination of vent stacks, bollards or ground level boxes, depending on the gas regime and wind conditions at a particular site. A schematic layout for the barrier is shown in Figure 1.

The system is particularly effective where gas migration is occurring through shallow layers of sand and gravel up to 5m depth, underlain by an impermeable layer. This is typical of many situations encountered in the UK. The nodes can be installed to a maximum depth of 5m below starting level. The starting level can be in trenches up to 3m depth, giving a maximum effective depth of 8m. As the depth of the migration pathway increases below the toe of the nodes the barrier becomes less effective.

(Click on image to see larger version)

Generally the flow of gas of gas in the ground towards a well or barrier can be modelled using the equations for planar flow of fluids based on Darcy's law. One of the most common situations is shown in the conceptual; model in Figure 2, where a permeable layer is overlain by an impermeable barrier (a capping layer or hard cover ). A relatively shallow groundwater table or impermeable clay layer typically provides a lower confinement to gaseous flow. In this situation we may use the equation for flow of fluids in a confined aquifer towards a horizontal slot based on work by Chapman (1959) relating to groundwater flow and the United States Environmental Protection Agency (1975) which looked at the performance of vent wells in landfills. This approach is also proposed by Oweis and Khera (1990).

When analysing gas flows in the ground the conceptual model of the migration pathways must be correct, otherwise the analytical results will be unrealistic. Other models may require the use of different flow equations, for example if the impermeable cover layer is absent.

Flow to a slot in a confined aquifer is given by

(1)

(Based on Darcy flow), where:

Q = flow of ground gas in m³/s from one side of barrier, over length of barrier, L,

K_i = intrinsic permeability of the ground in m².

g = unit weight of landfill gas in N/m³

µ = viscosity of gas being considered in Ns/m²

T = thickness of confined aquifer or migration layer in m

L = length of section of barrier being considered in m

D P_g = driving pressure of gas from ground as m head of water (gas pressure/unit weight)

L_o = distance of influence of barrier in m

This estimated gas flow from the ground is the volume that requires dilution in the duct.

The calculated flow of gas towards the barrier is very sensitive to the chosen value of the distance of influence of the barrier and the permeability of the ground. Evidence^{, ,} suggests the radius of influence for passive vent wells is most likely to vary between 1m and 10m, depending on ground conditions, type of well, etc, and it seems reasonable to use similar values for L_o in this case.

However because the calculated value of gas flow is sensitive to any variation in L_o and permeability, a sensitivity analysis should usually be carried out. It is also important to note that the actual flows are affected by many other factors, some of which are dynamic (eg changes in atmospheric pressure, temperature, gas compression), and therefore difficult to allow for except in complex mathematical models. For this reason the results from the simplified analysis method proposed should not be considered as absolute, but rather as an aid to assess the effects of varying layouts and to support engineering judgement.

Using these equations and the measured pressures from monitoring wells, the flow of gas to the line of vent wells can be estimated. When a gas monitoring borehole is left in the ground with the tap closed, the atmosphere within it comes into equilibrium with the surrounding ground, and on opening a peak value of pressure is recorded as the well comes into equilibrium with atmospheric pressure. This represents the ambient pressure in the surrounding ground, which is considered as the driving pressure for gas in the ground towards the vent nodes.

At the top of the vent node the differential pressure is approximated to atmospheric pressure less the reduction due to flow of air through the dilution duct⁹.

The flow capacity of a single geocomposite vent can be calculated directly using Darcy's law, and the value of intrinsic permeability, K_i, for the particular geocomposite used.

The flow in the vents is given by;

Total flow capacity of vents (2)

Where

K_i = intrinsic permeability of geocomposite in m²

A = area of vents in m²

N = number of vents

i = pressure gradient along vent node = D P_v (in m head of water)/length of vent node.

The sum of the flows from all the vents must be greater than the flow into the system from the surrounding ground. An alternative check is to assume that the driving pressure in the ground reduces to zero at the top of the node. The sum of the permeability and the flow area of the vents per metre length of barrier can then be simply compared with that of the ground, such that the following condition must apply to ensure the vents can provide a preferential flow path for gas:

K_{i vent} x A _vent > K_{i ground} x A _ground (3)

The flow of fresh air through the collection/dilution duct, required to dilute the methane to less than 1%, can be calculated using the gas flow calculated in (1) and the guidance in CIRIA Report 149 and British Standard BS 5925.

Using: from CIRIA 149 (4)

Where c_e = design equilibrium concentration in duct and at outlet vents, %

Q_duct = fresh air flow through each duct length, L, in m³/hr

Q = flow of ground gas into the system for each duct length, L, in m³/hr

The ventilation area required to provide this flow can be calculated using the guidance for designing natural ventilation provided in British Standard BS 5925. Using these design criteria the arrangement and type of ventilation can be determined.

There will be some friction losses in the duct, but considering the very low flow rates that occur in passive ventilation systems, these are considered to have negligible effect. In any event they can be allowed for by increasing the amount of ventilation provided.

The calculations require a factor of safety to be incorporated to allow for the effects of:

• uncertainty in the gas regime,
• friction losses due to constrictions to flow in the system,
• blocking of vents or other breakdowns of the system.

It is usual to apply the following factors of safety in gas ventilation design

• the use of maximum concentrations, flow rates and pressures regardless of spacial or temporal variation across a site often gives an inherent factor of safety, because the calculations assume constant flows from the ground across the whole site, at the design values,
• design gas values - apply a factor of safety of between l and 5 depending on the amount and reliability of the gas monitoring data and site investigation data,
• on overall ventilation air flow - apply a factor of safety of between l and 5 depending on the sensitivity of the development , risk, what management systems will be in place, how critical the dilution barrier is, etc,

• on ventilation outlets - apply factor of safety of between l and 3 to the calculated vent area on the same basis as the air flow.

The passive dilution barrier is installed using a unique no dig method in which a steel mandrel is vibrated up to 5m into the ground, using a vibrating piling hammer supported by a 360^o excavator. (Figure 4). Once the hollow mandrel is in the ground the central cutting shoe can be removed (see Figure 5) and a geocomposite strip inserted. The mandrel is then withdrawn, leaving the vent in the ground.

Figure 4 vibrating mandrill into ground

The key advantages of this method of installation are:

• speed – up to 30 vents per day can be installed,
• cost – there is a reduction in excavation costs and disposal of spoil that is frequently contaminated,
• safety – contact with contaminated materials by the installers is minimised.

A further advantage is that walls can be constructed very close to site boundaries and in areas where access is restricted and conventional barriers could not be constructed, as shown in Figure 4.

A site trial of the new system was undertaken at a landfill site in North West England. The site was formally a brickworks which ceased operations in 1975, leaving open clay pits. Filling of the site began in 1981, with approximately 2.5 million tonnes of domestic refuse being placed. The site was completed in 1995 leaving depths of waste up to 33m, which was covered by a capping layer.

The site is underlain by Glacial Till overlying Millstone Grit and the Till generally comprises relatively impermeable clays which act as a natural barrier to landfill gas migration. The site has been retro fitted with a gas extraction system which collects the gas and burns it off at flares. Routine monitoring by the landfill operator and the Environment Agency identified one area where gas appeared to be migrating off site. The monitoring borehole in question was approximately 20m outside the landfill, beyond the influence of the extraction system.

The migration is thought to be occurring along a granular lense or infilled glacial overflow channel within the Glacial Till, which comprises sand and gravel. These features are common in the area. The conceptual gas migration model is shown in Figure 6.

Before installation of the barrier methane concentrations in one monitoring borehole were consistently in excess of 30% v/v with peak levels of 50% v/v. Carbon dioxide concentrations were typically between 10% v/v and 20% v/v.

(Click on picture to view larger version)

The passive barrier was installed over a length of 20m, offset from the affected borehole by 1m. It runs 10m either side of the borehole, between it and the centre of the landfill. It is 75m from the nearest extraction well within the landfill and 20m from the landfill boundary.

The passive dilution barrier comprises 14 No vertical geocomposite vent nodes (410mm by 30mm) spaced at 1400mm centres. They are driven to a depth of 5m below ground level. A collection/dilution duct has been placed over the nodes and is 450mm deep by 410mm wide. It is vented via a 3m vent stack at one end with a 0.9m high inlet bollard at the opposite end, both 150mm diameter.

The system was installed over a period of 4 days and was commissioned on 18 October 2000.

Gas monitoring has been undertaken on a daily basis before and after installation of the barrier. The results presented in Figures 7 and 8 show a clear reduction in gas concentrations in the borehole where migration was occurring, after the barrier was commissioned. Both methane and carbon dioxide concentrations have dropped to generally less than 1% v/v in the ground. This demonstrates the effectiveness of the system. It is particularly effective where the vent nodes penetrate the full depth of the migration layer, although evidence does suggest the effectiveness reduces as depth of the migration pathway increases below the toe of the nodes. Currently it is possible to install the nodes to maximum depths of around 8m below ground level, if a 3m starter trench is excavated.

Click here to see graph

The gas monitoring has been undertaken over a range of atmospheric pressure conditions (Figure 9) with a consistent reduction in gas being recorded in the borehole where migration was occurring.

Click here to see graph

Click here to see graph

The passive dilution gas migration barrier offers several advantages over conventional vent trenches and vent wells:

• speed of installation,
• reduced costs,
• increased safety as contact with contaminated materials by the installers is minimised,

• efficient ventilation dilutes gas emissions to tolerable levels,
• can be installed in restricted areas.

The system can be designed to deal with different ground conditions, gas regimes and wind conditions to ensure the safe venting of gases at tolerable concentrations, using accepted principles of fluid flow in the ground. A monitored site trial has demonstrated the effectiveness of the barrier in reducing landfill gas migration. Further trials are planned to confirm the distance of influence of the system and to measure air flows through the dilution duct.

The authors are grateful to SEL Environmental Limited for permission to publish this paper. They would also like to thank Gavin Davies and Jake Hacker at Arup Research and Development for their invaluable discussions. The concept and installation method used for the vent nodes is subject to patent.