There is no escaping the fact that nuclear power plants are very expensive to build relative to all other forms of electricity production. Actual and projected costs vary depending upon a range of factors, but the front-loaded cost structure (high initial investment then relatively low running cost) according to a recent report from MIT assumes that a nuclear reactor costs $4000 per kilowatt of electricity produced to build — or $4 billion for a typical one — gigawatt nuclear power plant. Actual industry estimates for reactors being built today are at least $6 billion and as much as $10 billion. ‘‘If you build a nuclear power plant and operate it well, it’s going to produce a steady stream of income’’, but ‘‘the disadvantage of nuclear is the enormous capital commitment that is made up front’’.105

As the MIT report goes on to say, ‘‘the track record for the construction costs of nuclear plants completed in the USA during the 1980s and early 1990s was poor. Actual costs were far higher than had been projected… The first few US plants will be a critical test for all parties involved’’.106

However, ‘‘new nuclear plants are designed to last longer without major upgrades — up to sixty years, vs. thirty to forty years for gas, coal, and most renewables. And they can generate as much or more power than the largest fossil plants, up to 1.6 GW or more, while saving 7-8 million tonnes a year of carbon emissions’’.107 Due to their capital intensity, long construction times and proclivity for cost overruns, nuclear reactors are extremely expensive to build and are plagued by uncertainties for investors. xlii

xliThis led RBC Capital to argue that uranium spot prices will double by 2012. China recently entered into a contract with the French firm, AREVA, which will see the French firm sell $3.5 billion worth of uranium to China.

хі11 Nuclear also has the additional burden of having to set aside funds directly, or in some countries via a government imposed levy on electricity production, in order to pay for the costs of waste disposal and decommissioning of the plant.

The financial risks of new nuclear reactors has led Wall Street firms to indicate that it will be difficult to sell bonds to support these projects in the capital markets. Even in the United States, which under its Energy Policy Act of 2005xl111 authorized $18.5 billion in loan guarantees to cover 80% of building the first tranche of new reactors, the incentives have not been enough to catalyze significant private sector interest investment thus far.109 The market remembers the Washington Public Power Supply System fiasco, when in 1983 it defaulted on $2 billion worth of bonds as a result of mothballing four power plants after a string of delays and cost overruns. According to a major engineering organisation, UK utilities have cut back their nuclear build connection plans by 28% to 18.4GW, with a possibility that this may further reduce to only 13 GW when final decisions are taken. Furthermore, industry requires some surety that government will remain committed to a nuclear future. Such concerns are exacerbated by ‘‘major uncertainties surrounding individual site planning, grid connection, licen­sing, waste and finance which need to be resolved without further delay and excessive consultation’’.110,xliv

It is clear that in a fully liberalised energy market with government setting no targets or market incentives, utilities would not build nuclear power plants. This is why carbon trading is extremely important for the future economic viability of nuclear power because it pushes up the cost of its two main rivals: coal — and gas-fired stations. Without a price being attached to carbon, via the European Emission Trading Scheme (ETS) or some other mechanism, coal and gas are the cheapest options with nuclear in third place. However, if carbon were to be 25 euros tonne 1 then nuclear becomes the cheapest. One of the measures in the UK government’s reshaping of the energy market is precisely to impose a ‘‘floor’’ price on carbon.112

The projected versus actual costs of reactors do not make for good reading. One assessment explored estimated and actual US nuclear power plant construction costs from 1966 to 1977 (when the majority of American reactors were built) and found that in every case plants cost at least twice as much as expected.113,114 The last reactor to be built in the UK, Sizewell B, cost £1.8 billion as opposed to a projected £300 million, and the new EPR reactors being built in Finland and France are both over time and budget. Assessing costs in

xliiiMoreover, the 2005 Energy Act also streamlined the regulatory approvals process given the long delays in previous builds. Under the previous regime, once a plant had been constructed objections could be raised. The law also allows for a ‘‘risk insurance’’ to protect utilities against unexpected federal or state regulatory delays for up to six new reactors built under the new licensing structure. The loan guarantees provide 80% of the project cost to be repaid over 30 years. However, these loan guarantees are not having the desired effect, witness the recent decision by Constellation Energy Group Inc to pull out of a joint venture with EDF to build new nuclear power plants, citing costs of government loan guarantees to build the plants. This highlights the complex and at times unfavourable economics in a liberalised energy market even with government subsidies.108

xliv ‘‘The Government intimated that, consistent with its non interventionist free market approach, it would play its part as a market ‘enabler’ by reducing or removing any barriers impeding the new build programme’’.111

Asia and Russian is much more difficult given that they do not have liberalised energy markets.115 However, concerns over the negative PR engendered by the delays in France and Finland have led some of the major plant construction companies to take out full page adverts in the leading nuclear industry maga­zines, such as Nuclear Power International Magazine, emphasising the claim that in China new nuclear plants are being built on time and on budget. xlv In the USA whilst a small number of proposals for new plants are proceeding, an equal number have collapsed due to some partners in the consortia deciding the numbers do not stack up, even with the existence of federal loan guarantees. Indeed, Florida Power & Light (FPL) has suspended work on a new two reactor project citing the “deteriorating regulatory environment’’.116

One of the challenges that corporations face when building new plants, is what the industry itself calls ‘‘first-of-a-kind engineering’’ (FOAKE) challenges. The areas of greatest uncertainty are the design and construction, financial backing, the political environment and severe weather during construction, because it is difficult to estimate their total overall impact to the project and to effectuate reasonable control. Capital cost of a ‘‘nth-of-a-kind’’ (NOAK) reactor should be 10 to 20% less than that of a first-of-a-kind (FOAK) reactor because of the lessons learnt in the construction and deployment of earlier units.117 Moreover the “manufacturing learning curve generally flattens out after five to seven repeat units have been built…’’, even so nuclear reactors are expensive to build and vulnerable to fluctuations in national interest rates. As other countries have demonstrated (e. g. France and Japan), the cost structure of nuclear is such that only a large new build programme can generate ‘‘sufficient economies of scale to compete with gas-fired generation technology in the absence of effective carbon trading. In one sense, nuclear power is an all-or-nothing option’’.118

In Asia, current real-world costs are significantly lower than in North America and Europe. The two leading reactor designs now being built in China are the indigenous CPR-1000 and the Westinghouse AP-1000. Reported capital costs are in the range of $1296-$1790kW 1 (ref. 119). Korea has focused attention on its APR-1400 design, with domestic overnight costs of $2333 kW 1 (WNA 2010d) A recent contract for $20.4 billion has been signed with Korean consortium KEPCO to build four APR-1400 reactors in the United Arab Emirates (UAE), at a turnkey cost of $3643 kW 1. This price is notable con­sidering that it is offered under near-FOAK conditions, because these will be the UAE’s first nuclear plants.

There is a great deal of regional variability then in the capital cost of building nuclear reactors. A significant part of the cost is wages for the large number of skilled and unskilled workers necessary to build and operate the nuclear plant. In addition, there are concerns that there are not enough skilled workers for an expanded new build programme.

xlv ‘‘In China, four new AP1000s are currently under construction and they are being built in an on — time and on-budget manner, with the first scheduled to come online as planned in 2013’’. (Nuclear Power International Magazine, 2010).

The financial effects of an increase in insurance for the viability of new nuclear plants in light of Fukushima are at present unknown however one recent review suggests that ‘‘it is hard to see how a regulatory body could give generic approval to any new design until the problems at Fukushima have been thoroughly understood and designers find ways to prevent a repeat of events. As a result, completion of the U. S. and U. K. generic reviews will be delayed. The extent of additional design requirements remains to be seen, but it is very likely that additional costs, perhaps significant, will be imposed on any new designs’’.171

There appears to be little information (at least in the English language) in the scientific literature on the social and psychological impacts of Kyshtym on the population. A report by Collins15 highlights the delays in evacuation of many contaminated areas and the lack of information, and humanitarian and medical aid given to exposed populations. This contributed, unsurprisingly, to strong anti-nuclear feelings amongst the Kyshtym population termed by Collins15 a “psychological animosity’’ towards nuclear power.

Intermediate level wastes are ‘‘materials with radioactivity levels that exceed the upper limits for LLW but do not need heating to be taken into account in the design of storage or disposal facilities’’.2 The majority of ILW arises during spent fuel reprocessing at the Sellafield site and consists of heterogeneous materials such as: magnesium alloy cladding that is stripped from Magnox fuels; steel, zircaloy and graphite components from AGR and PWR fuels; aqueous waste, sludges, flocs and organic materials from PUREX and radionuclide waste stream purification treatments; aqueous waste, flocs and filters from pond water treatments; and contaminated machinery from reprocessing operations (see Figure 1).2 Most of the remaining ILW is produced at nuclear power sta­tions where ILW principally arises during reactor operations and during sub­sequent decommissioning. Historical waste storage activities at a range of nuclear sites also give rise to ILW. As a consequence, the ILW waste stream has a highly heterogeneous and chemically challenging composition (see Figure 1).2

Animals can be contaminated through the skin, by inhalation, and, most importantly, via ingestion of radionuclides. Uptake through the skin is not
usually an important route of contamination, and is not considered here. Inhalation by animals is potentially more important than skin absorption since the lung surfaces, the site of gaseous exchange, are more permeable to a wider range of elements. Radionuclides may be inhaled in different forms, including gaseous compounds, aerosols and particles by terrestrial animals. The ability of radionuclides to pass through the pulmonary membranes varies considerably. Despite low transfer rates for actinides, such as plutonium, they are often more readily absorbed via the lungs than the gastrointestinal tract. Gaseous iodine is readily absorbed and inhalation may have been a route of contamination of milk of housed animals following the Chernobyl accident and currently in Japan.67 Inhalation, however, is generally not a major contamination route for most radionuclides for animals and is not considered further here.

The most important transfer pathway to animals is the ingestion of con­taminated food, soil and drinking water. Intake via drinking water is generally a small contributor to total radionuclide intake. Radionuclide intake via soil can be significant, but the availability for absorption of soil-associated radio­nuclides may be lower than plant incorporated radionuclides, although there is only evidence of such a difference for caesium.29 Hence, it is the ingestion of contaminated feed and processes influencing absorption and retention that usually determines the radionuclide content of animals.

As any fuel is irradiated, the proportion of useful fissile isotopes decreases and the content of fission products increases. Some of the fission products are efficient neutron absorbers, and ‘‘poison’’ the fuel. Consequently, after approximately three years irradiation, the reactivity of the fuel is too low and it is necessary to remove the ‘‘spent’’ fuel from the reactor and replace it with fresh. The composition of a typical uranium fuel is summarised in Table 2.

The Rocky Flats Environmental Technology Site, formerly Rocky Flats Nuclear Weapons Plant, is located northwest of Denver, Colorado, and between the years 1952 and 1989 was responsible for the production of com­ponents for the United States nuclear weapon program. This involved the use of various radioactive materials such as plutonium and uranium as well as toxic metals and hazardous solvents. Two main events are responsible for the release of plutonium outside the Rocky Flats Plant boundaries. These events were a fire that occurred in the plutonium processing building in 1957 and wind-blown releases occurring mainly during 1968 and 1969 from an outdoor waste storage area called the 903 Area. An estimated 5000 gallons of plutonium — contaminated waste leaked from the waste containers covering an area of 22 500 m2 according to monitoring conducted in 1968. Windstorms in 1968 and 1969 blew the plutonium-contaminated soil off the site thus contaminating a much greater area, with an estimated 66.6 to 518 GBq of 239+240Pu released to the off-site environment.30

Soil samples analysed from a series of pits along a contaminated topose — quence at Rocky Flats revealed plutonium contamination ranging from 2220 to 11460 Bq kg-1, with a mean activity of 7250 Bq kg-1, and 241Am contamination ranging from 1840 to 8840 Bqkg-1, with a mean activity of 5480 Bqkg-1.31 The activity was primarily located in the uppermost layer of the soil with 90% of the contaminants distributed in the top 20 cm in four of the five pits.

Synchrotron radiation studies revealed the oxidation state of the plutonium in the soils and concrete as Pu(IV) and identified its chemical form as the insoluble hydrous oxide PuO2 • xH2O (ref. 32) with transport of the plutonium likely confined to fine particle migration.

As discussed, initial planning for UK GDF implementation is subject to sig­nificant uncertainty and at this stage flexibility in approach is essential. Parti­cularly, a range of potential host geologies and repository designs must be considered at a generic level and any advanced case studies are presumptive and thus inappropriate at the current pre-site selection stage. Consequently, the NDA-RWMD has implemented a preparedness approach and has recently published a document that outlines the current UK GDF implementation strategy.14 In this paper, a matrix of generic host settings reflecting typical, potentially suitable UK geologies and repository designs have been selected to demonstrate the viability of UK geodisposal, and better inform the conceptual design processes. The resulting host rocks where the GDF may be located include: higher strength rocks (typically crystalline igneous, metamorphic, or geologically old sedimentary rocks where fluid movement is supported through rock fractures e. g., granite); lower strength sedimentary rocks (typically young sedimentary rocks where fluid movement is through pore spaces e. g., clay); and evaporites (rocks that result from the evaporation of water containing salts e. g., halite). Further, the covering rocks included in the generic geological assess­ment include host rocks to the surface, or sedimentary cover rocks. The resulting matrix of possible geological scenarios (see Table 3) indicates that all of the above potential geological combinations are possible in the UK, with the exception that that UK evaporite deposits do not extend to the surface.

Reflecting the geological considerations, the NDA have identified a matrix of illustrative GDF concepts (see Table 4) to inform their scoping work. These concepts reflect international experience (see Table 1), but also consider the existing UK ILW geodisposal concept. This relatively well developed concept19 stems from the UK’s unsuccessful attempt to implement geodisposal in the 1990s (see section 3.3.1). In developing this concept, it was assumed the GDF

would be housed at several hundred metres depth in a geological setting that comprised high permeability sedimentary rocks overlying low permeability sedimentary rocks, with the GDF housed in low permeability hard rock (see Hicks 2008; ref. 11). In this model, after conditioning (in cement and steel), the waste would be stored then emplaced in a GDF as the facility becomes avail­able. After several decades to centuries, when waste emplacement had occurred and when the operational lifetime of the GDF was complete, the GDF would be sealed with a cementitous backfill and abandoned. At this point, the sub-surface GDF environment would then contain conditioned and packaged waste, along with significant volumes of cement and structural iron, and the sub-surface would resaturate with groundwater. The cementitous waste packaging and backfill were designed on resaturation to maintain high pH conditions pro­moting hydrolysis of metal ions (including the actinides) and thus minimising radionuclide solubility. In addition, the use of iron metal in the facility (in both packaging and engineering structures) is intended to promote strongly reducing conditions, again limiting the solubility of some radionuclides.

JAN PENTREATH

ABSTRACT

Radiological protection has a long pedigree; its origins go back almost a century. Since then, not only has a large body of information on the effects of radiation been accrued, but this information has been used in a most successful way to manage the exposures of people in all forms of exposure situations. This success has largely been due to a unique orga­nization, the International Commission on Radiological Protection (ICRP), that has continually evaluated, interpreted and worked out how best to apply the knowledge that has arisen from the scientific disciplines of radiation physics, dosimetry and radiobiology, together with the complicated interpretation of numerous epidemiological studies. Fur­thermore, the ICRP has attempted to interface the continually improving science with the ever changing cultural and sociological context within which radiological protection needs to be applied. This is an on-going task. The current situation is one in which radiological protection gui­dance is set out within a framework of three exposure situations (planned, emergency and existing), involving three categories of human exposure (medical, occupational and public). Due to the scientific interpretation of the data, this matrix of exposure situations, and categorization of those likely to be exposed, is handled within a set of principles of justification of exposure, the optimization of the level of protection and the application of dose limits. All of the elements of this framework are briefly set out and discussed in this chapter, together with a brief overview of the current rates of exposure, due to different exposure situations, for people within the UK.

Issues in Environmental Science and Technology, 32 Nuclear Power and the Environment Edited by R. E. Hester and R. M. Harrison © Royal Society of Chemistry 2011

Published by the Royal Society of Chemistry, www. rsc. org

200

1 Introduction

There is probably more known about the effects of radiation on human beings than there is about the effects of any other hazard relating to the generation of energy on a large scale. In addition, this knowledge is better organized, man­aged and converted into useful practical procedures at an internationally agreed level, than anything comparable in relation to other hazards. The reasons for this lie in the fact that an enormous data base has been accrued over almost a century, and because of the establishment of what is a unique and, by now, a somewhat ancient body — the International Commission on Radi­ological Protection (ICRP).

Its origins lie in the fact that, following their discovery at the end of the 19th century, the medical benefits of X-rays, and of the gamma rays from radium, were very quickly recognized. The dangers of radiation soon became apparent, however, and national committees set up to address the problems started to appear in 1913. There was also a need for some form of international co-operation, and following a decision by the Second International Congress of Radiology, the ICRP was established in 1928 under the name of the Interna­tional X-Ray and Radium Protection Committee. It was restructured in 1950 and given its present name.

The ICRP is an advisory body. It regularly issues detailed advice and information regarding protection against the hazards of ionizing radiation and, at suitable intervals, revises its overall set of “Recommendations”. The first report in the current publication series contained the Recommenda­tions that had been adopted in 19581 and, since then, revisions have been set out in Publication 26,2 Publication 603 and, most recently, in Pub­lication 103.4

The advice of the ICRP is aimed principally at regulatory authorities, organizations, and individuals that have responsibility for radiological protection, and virtually all international standards and national regulations addressing radiological protection are based on its recommendations. There is a close connection between the Recommendations and the International Basic Safety Standards for Protection against Ionizing Radiation and the Safety of Radiation Sources (usually simply called ‘‘the BSS’’), which are co-sponsored by the relevant international organisations within the UN family and issued by the International Atomic Energy Agency (IAEA). The governing body of the IAEA requires that the BSS take the ICRP’s Recommendations into account. These Recommendations are then, in turn, cascaded down to such bodies as the OECD’s Nuclear Energy Agency (NEA), and to regional bodies (such as EURATOM), and to national bodies, such as what was the UK’s National Radiological Protection Board, the functions of which are now part of the Health Protection Agency (HPA).

The ICRP operates via a set of five committees, each of which also makes extensive use of specialized task groups. It also works closely with the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) that was established in 1955 by the UN General Assembly with a mandate to

assess and report on levels of, and exposure to, ionizing radiation. The ICRP thus draws upon a vast amount of world-wide experience and data bases on all forms of radiation exposure relating to medical practices, industrial workers and the general public, and, more recently, on the natural environment as well.

The first dose unit, the roentgen (R), was established for X-rays in 1928 by yet another committee, the International X-ray Unit Committee, which was later to become the International Commission on Radiation Units (ICRU). The first official use of the term ‘‘dose’’, together with an amended definition of the unit R, appeared in the 1937 recommendations of the ICRU.5 The ICRU raised the concept of an absorbed dose, and officially defined the name and its unit, the ‘‘rad’’, in 1953 to extend the concept of dose to certain materials other than air.6 The first dose quantity incorporating the concept of different responses of tissues to different types of radiation (known as ‘‘relative bio­logical effectiveness”, or RBE), was the ‘‘RBE dose in rems’’ (rem stood for roentgen equivalent in man, or mammal). This was an RBE-weighted sum of absorbed dose in rads, proscribed in the 1956 recommendations of the ICRU. As a result of joint efforts between the ICRU and the ICRP, it was later replaced by the ‘‘dose equivalent’’ which was defined by the product of absorbed dose, the quality factor of the radiation, the dose distribution factor and other necessary modifying factors.7 The ‘‘rem’’ was retained as the unit of dose equivalent. The ICRU also defined another dose quantity, ‘‘kerma’’, and changed the name of ‘‘exposure dose’’ to the simple one of‘‘exposure’’ in its 1962 recommendations.

Since then much has happened, including the introduction of SI units. The fundamental dosimetric quantity in radiological protection is now the absorbed dose (D). This is the energy absorbed per unit mass, and its unit is the joule per kilogram, which is given the special name gray (Gy). Absorbed dose is defined in terms that allow it to be specified at a point, but it is used by the ICRP, except where otherwise stated, to mean the average dose over a tissue or organ. Multiplying the absorbed dose by appropriate weighting factors, depending on the type of radiation, creates the equivalent dose (HT) in the relevant organ or tissue. The equivalent dose is preferred in radiation protection because it is more closely related to the risk of harm in the exposed organ or tissue. By weighting the equivalent dose in each organ in proportion to its radiation sensitivity (in other words, to the probability and severity of the harm done by radiation), and then adding the weighted contributions from each organ to a total body dose, a third dose, the effective dose (E), is obtained. In radiation protection it is usually the effective dose that is determined for comparison with dose limits or for assessments of risks. Both the equivalent dose and the effective dose are measured in a unit called the sievert (Sv). For some appli­cations, a collective dose may be calculated, being the product of the number of exposed individuals and their average dose. The collective effective dose (S) may also sometimes be used as a measure of the expected collective harm, sometimes referred to as the ‘‘radiation health detriment’’.

Thus a system has been developed where the science base relating exposure to dose, and dose to effects, is examined, re-examined and interpreted by way of a set of conceptual and numeric ‘‘model’’ (see Figure 1). This modelling process

started with the creation of a ‘‘Reference Man’’, who has now evolved into a Reference Individual (male and female), and a Reference Person. The former is an idealised male or female entity with reference anatomical and physiological characteristics, as defined by ICRP.8 Thanks to the rapid development of the relevant technology, phantoms based on medical tomographic images, con­sisting of three-dimensional volume pixels (voxels), are now used to compute the mean absorbed dose in an organ or tissue, and these doses are multiplied by radiation weighting factors to provide the equivalent doses in the Reference Male and Reference Female. These steps are shown in Figure 1.

For the purposes of radiological protection, however, it is currently thought useful to apply a single value of effective dose irrespective of sex. This is achieved by deriving sex-averaged organ or tissue equivalent doses for an idealised Reference Person and then multiplying them by the corresponding tissue weighting factors.

The UK has a complex and diverse inventory of radioactive wastes described in some detail by Defra/NDA8 and the associated documents. Waste is classified primarily according to its radioactivity content. Low Level Waste (LLW) is waste with a radioactive content below defined levels (4GBq tonne 1 a-emitters, or 12GBq tonne 1 p — and p, g-emitters). Intermediate Level Waste (ILW) exceeds the threshold activity for LLW but is not so radioactive that it requires active cooling. High level waste (HLW) is the intensely radioactive fission product stream derived from fuel reprocessing and requires constant cooling due to the decay heat it generates.

In addition, the UK has materials which may be declared wastes in the future, and will then need to be managed accordingly. These include separated plutonium, where a small proportion is unfit for reuse in fuel and will have to be disposed as waste. Government is presently considering management options

for the rest of the plutonium stockpile, and one option is to declare it all as waste. The UK also holds large quantities of uranium (depleted, natural and reprocessed) in various forms which may be declared as waste. Finally, it is planned to dispose of a proportion of UK spent fuel as waste without repro­cessing. Current plans also assume that any spent fuel from new build reactors will eventually be disposed as waste. Collectively, ILW, HLW, plutonium, uranium and spent fuel are often referred to as ‘‘Higher Activity Wastes’’ (HAW). The UK inventory of wastes and potential waste materials, excluding new build waste, is summarized in Table 5, and options for their management are discussed in Chapter 5.

As was discussed previously, a number of contamination issues exist at the Y-12 complex at Oak Ridge. In order to address these concerns, two permeable iron reactive barriers were installed at the Y-12 plant, consisting of two pathways, in 1997.113 As water flows through the barrier, the reactive medium (in this case Fe0) traps or degrades the contaminant. Pathway 1 at the Y-12 complex was designed to capture groundwater in a gravel-filled, high-density-polyethylene — lined trench. The groundwater would then be treated within a vault containing zero-valent iron. The second pathway involved a permeable trench, in a sub­parallel direction to groundwater flow. The trench, 2 ft wide and 225 ft long, contained a 26 ft long zone of ZVI covered either side by gravel backfilled zones. Groundwater samples from monitoring wells both in, and downgradient, of the iron barrier at Pathway 2 contained only very low concentrations of uranium (<0.05mgl-1) compared to values found in groundwater samples in upgradient wells. This would suggest that ZVI is effective at immobilising uranium present in groundwater.114 Uranium concentrations in middle and deep wells located within the iron barrier displayed slightly higher than expected levels of uranium at ~ 0.2 to 1 mgl_1. These wells are located in the upgradient portion of the iron barrier where upward hydraulic gradients dominate. The higher concentrations seen here may therefore be a result of a higher inflow of untreated groundwater. Some downgradient wells also showed higher than expected uranium concentrations suggesting that treated groundwater is being re-contaminated from the mobili­sation of uranium on downgradient soils or that groundwater flows not treated by the barrier are reaching the wells.114

Another contrasting field research study was performed using reduction of U(vi) to U(iv) as a method for immobilising the contaminant. Subsurface conditions favourable for bioremediation were established38 followed by peri­odic injection of ethanol. The pH of the test area was adjusted to pH 5-6 causing an increase in uranium sorption and resultant decrease in groundwater uranium concentrations from ~300 (at an initial pH of ~3.4) to ~5 pM.115 Ethanol injections began on day 137 and ended on day 535. Following on from an initial denitrification phase (day 137-184), a period of uranium and sulfate reduction occurred (day 184-535) during which uranium concentrations in the groundwater decreased from 5 to 1 pM.115 XANES analysis confirmed that between 39% and 53% of the uranium recovered from the sediments after biostimulation was reduced U(iv).115 The results from this study, where U(vi) reduction correlated with sulfate reduction, contrast to those from a similar study at the Rifle Site (discussed previously), where aqueous uranium con­centrations rebounded when sulfate reducing conditions became dominant.28 It is possible that sulfate-reducing bacteria that are capable of U(vi) reduction were stimulated by the use of an electron donor (ethanol) in this study, in

contrast to the use of acetate at the Rifle site which may have stimulated alternative organisms. Further studies at this site are required to assess the long-term resistance of U(iv) to re-oxidation and remobilisation.

A recent column-flow experiment demonstrated the potential for the remedia­tion of uranium and technetium in low pH, highly contaminated environments, such as Oak Ridge (as discussed previously) through co-precipitation. Luo et al. showed that in conditions found at Oak Ridge, greater than 95% of soluble uranium and 83% of technetium can be co-precipitated with Al-oxyhydroxides by raising the pH above 4.5 with the addition of a strong base (NaOH).39 The precipitated uranium and technetium were found to be stable in the presence of high nitrate concentrations [50 mMCa(NO3)2] and low carbonate concentrations.39