Thermal conductivity and diffusivity estimations for shallow geothermal systems

Busby, Jon. 2016 Thermal conductivity and diffusivity estimations for shallow geothermal systems. Quarterly Journal of Engineering Geology and Hydrogeology, 49 (2). 138-146. 10.1144/qjegh2015-079

Before downloading, please read NORA policies.
revision submitted Feb 2016.pdf - Accepted Version

Download (911kB) | Preview


Horizontal closed-loop ground collectors for ground source heat pumps are located within the soil and the top of the underlying superficial deposits. Estimating thermal properties for this zone is difficult as it is heterogeneous and is subject to seasonal water content variations. Soil thermal diffusivity values have been calculated at 56 sites using temperature data from UK Met Office weather stations. The technique utilizes the decrease in amplitude and increase in phase shift with depth of a transmitted heat pulse in the ground, the magnitudes of which are determined by thermal diffusivity. The weather stations are located throughout Great Britain and incorporate different soil types. The apparent thermal diffusivities derived from seasonal temperature cycles spanning several years generate seasonally averaged site-specific estimates that can be considered alongside diffusivity values determined in the laboratory or obtained by point measurements using field needle probes. Associated thermal conductivities have been estimated from the thermal diffusivities from knowledge of soil texture. Median thermal conductivities for the sand, loam and clay soil types have been estimated as 1.56, 1.15 and 1.81 W m−1 K−1 respectively with corresponding thermal diffusivities of 0.9961 × 10−6, 0.7173 × 10−6 and 1.0295 × 10−6 m2 s−1 respectively. Shallow ground source heat collector loops often comprise straight pipes or coiled pipes (commonly referred to as slinkiesTM) that are laid horizontally along the base of a trench, or coiled pipes that are inserted vertically in a slit trench (Banks 2012). The suggested depth of the trenches varies, but GSHPA (2014) recommended 0.8–1.5 m below ground level and Banks (2012) indicated 1.2–2 m. These trenches are therefore located within the soil and the top of the underlying superficial deposits. This unconsolidated geological material is often referred to as the parent material of the soil and is a geological deposit over and within which a soil develops (Lawley 2008). Soils can be categorized as sand, silt, clay and loam (or combinations of these) where a loam is composed of approximately equal amounts of sand, silt and clay. The length of the collector loop depends on many factors, but the ground's thermal properties (thermal conductivity and thermal diffusivity) will need to be either estimated or measured (e.g. IGSHPA 1996; VDI 2001; Preene & Powrie 2009; Banks 2012; Curtis et al. 2013; GSHPA 2014) to ensure adequate sizing of the loop. There is a paucity of data on soil thermal properties required for the sizing of horizontal collector loops that is compounded by their seasonal dependence. A field method for estimating soil thermal properties has been given by IGSHPA (1989). Many quoted measured soil thermal properties are based on laboratory measurements (e.g. Clarke et al. 2008). These often use bulk soil samples that are bagged in the field, in which case the in situ consolidation is lost and is re-created in the laboratory. However, this will alter the bulk density, which is an important parameter in determining the thermal properties (e.g. Kersten 1949). Alternatively, field samples can be taken with a corer that incorporates a liner to preserve the natural texture and moisture, before transfer to the laboratory for thermal properties testing. For borehole-based vertical systems, a thermal response test can be performed to measure in situ bulk thermal conductivity (e.g. Banks et al. 2013), but there is at present no equivalent for horizontal systems. Thermal conductivities at a point on the ground can be measured with a needle probe (Campbell et al. 1991; Bristow et al. 1993; Bilskie et al. 1998). Field probes are mounted on a long handle so that they can be inserted into the base of auger holes to over 1 m depth. The probe generates a constant heat output and is a transient technique that monitors the increase of temperature with time. The determined thermal conductivity is representative of only a small cylindrical volume around the probe and errors can result from the contact between the probe and the soil. King et al. (2012) have indicated that a minimum of 12–16 measurements should be taken at a site with a field probe to produce a representative geometric mean thermal conductivity. However, such values are still valid for only a particular point in time, as near-surface thermal properties are affected by the seasonal variation in soil moisture. As an example of this variation, Gonzalez et al. (2012) quoted a 37% increase in soil thermal conductivity at 0.75 m depth and a 23% increase at 1 m depth between dry summer and wet winter conditions for a loamy sand (average composition clay 2.4%, silt 33.2%, sand 64.4%) that developed over a superficial deposit of sand and gravel. Apparent thermal diffusivity can be determined from soil temperature measurements and has been widely reported (e.g. Kappelmeyer & Haenel 1974; Adams et al. 1976; Horton et al. 1983; Verhoef et al. 1996; Gao et al. 2009). The technique utilizes the decrease in amplitude and delay in temperature change (phase shift) with depth of a transmitted heat signal applied to the ground surface, the magnitudes of which are determined by thermal diffusivity. If the heat signal is periodic (i.e. the diurnal or seasonal temperature cycle) and it is assumed that the heat transfer is governed by the 1D heat conduction equation, six different methods for calculating thermal diffusivity can be defined (Horton et al. 1983). Adams et al. (1976) and Horton et al. (1983) found that some of these methods gave erratic results. This may be partly due to using temperature measurements from the upper 10 cm of the soil, a zone where heat transfer is unlikely to be purely by conduction, and to too few temperature measurements, which do not adequately describe the periodic signal. This paper explores the calculation of soil thermal properties by utilizing the database of British meteorological soil temperature measurements taken at multiple depths to a maximum depth of 1 m. Thermal diffusivity is calculated directly from the depth-distributed soil temperatures and thermal conductivity is estimated from the diffusivity measurements with the addition of assumed parameters based on soil texture. The soil temperature measurements are widely dispersed covering many soil types and occupy the depth range of a horizontal ground collector loop. The calculated thermal properties are annual averages rather than a single seasonal value taken at a point in time. Although specifically incorporating British datasets the results and conclusions are applicable to shallow ground source heat in general.

Item Type: Publication - Article
Digital Object Identifier (DOI): 10.1144/qjegh2015-079
ISSN: 1470-9236
Date made live: 07 Jun 2016 13:56 +0 (UTC)

Actions (login required)

View Item View Item

Document Downloads

Downloads for past 30 days

Downloads per month over past year

More statistics for this item...