Table 2.

Estimated partition coefficients for ammonium reported for a selection of UK lithologies.

Lithology (and references)1Kd range (ml g−1)2Comments
Chalk (7, 10)0–0.03Low confidence as isotherms have not been identified
Triassic Sherwood Sandstone (1, 4, 5, 7, 8, 10, 12)0–0.2–0.6Reasonable confidence as there have been several independent tests, although most values are from one site only (Burntstump)
Lincolnshire Limestone (7)0.065–0.65Low confidence as isotherms have not been identified
Lower Greensand ‘Hassock’ (4, 10)0.18–1.8Low confidence as no isotherms have been found. Measurements have been on the ‘Hassock’ lithology only. A lower bound of zero should be used for Lower Greensand in general
Lower Greensand (undifferentiated) (4, 10)0–1.8
Red Crag (10)0.05–0.5Low confidence as isotherms have not been identified
Oxford Clay (7, 13)0.135–1.35Based on modelling of column experiments
Mercia Mudstone (7)0.5–5Low confidence as isotherms have not been identified
Gault Clay (7)0.65–6.5Low confidence as isotherms have not been identified
Lias Clay (2)1.2–2.6Low confidence as isotherms have not been identified. Cave & Taylor (2002) presented higher Kd* values for lower concentrations
Coal Measures Clay (13)0.18–1.8Based on modelling of column experiments, being derived from clay–sand mixtures. Value of 1.8 ml g−1 obtained for 100% clay system
Sand and gravel, clean (3, 6, 8, 11)0–0.4–0.9Reasonable confidence as there have been several independent tests. However, this is naturally a very heterogeneous lithology so a full site characterization should be made if the ‘clayey’ range of values is used
Sand and gravel, clayey (3, 6, 8, 11)0.4–0.9
Cohesive Boulder Clay (2) (Glacial Till)2–4Low confidence as isotherms have not been identified. Glacial Till is naturally a very heterogeneous lithology so a full site characterization should be made and the risk assessor must be very confident that there are no sandy sequences in the Till
Engineered clay landfill liners (9, 13)0.1–0.5–5Reasonable confidence as there have been a number of independent tests that yield seemingly consistent results. Kd has been shown to correlate with clay content and/or CEC so less conservatism might be permissible with suitable testing and justification. Consideration should be given to the dominant clay mineral present
  • References cited: (1) Butler et al. (2003); (2) Cave & Taylor (2002); (3) Ceazan et al. (1989); (4) Colley (1991); (5) Davison & Lerner (1998); (6) DeSimone et al. (1996); (7) Environment Agency (2000); (8) Erskine (2000); (9) Griffin et al. (1976); (10) Jackson (1989); (11) Kjeldsen & Christensen (1984); (12) Thornton et al. (2000); (13) Thornton et al. (2001).

  • 1Values are considered to apply equally to both the unsaturated and saturated zones.

  • 2Where two values for Kd are given this indicates that a uniform distribution is suggested for probabilistic modelling; where three are presented then a triangular distribution is suggested.