Global particle balance measurements in DIII-D H-mode discharges

https://doi.org/10.1016/j.jnucmat.2011.02.028Get rights and content

Abstract

Experiments are performed on the DIII-D tokamak to determine the retention rate in an all graphite first-wall tokamak. A time-dependent particle balance analysis shows a majority of the fuel retention occurs during the initial Ohmic and L-mode phase of discharges, with peak fuel retention rates typically ∼2 × 1021 D/s. The retention rate can be zero within the experimental uncertainties (<3 × 1020 D/s) during the later stationary phase of the discharge. In general, the retention inventory can decrease in the stationary phase by ∼20–30% from the initial start-up phase of the discharge. Particle inventories determined as a function of time in the discharge, using a “dynamic” particle balance analysis, agree with more accurate particle inventories directly measured after the discharge, termed “static” particle balance. Similarly, low stationary retention rates are found in discharges with heating from neutral-beams, which injects particles, and from electron cyclotron waves, which does not inject particles. Detailed analysis of the static and dynamic balance methods provide an estimate of the DIII-D global co-deposition rate of 0.6-1.2×1020 D/s. Dynamic particle balance is also performed on discharges with resonant magnetic perturbation ELM suppression and shows no additional retention during the ELM-suppressed phase of the discharge.

Introduction

Fuel retention, specifically hydrogenic isotope retention, in the plasma facing components (PFCs) of magnetic fusion plasma experiments is a research topic of interest mainly due to the potential for semi-permanent trapping of these isotopes in the PFC material. Retention mechanisms have been determined that can either saturate at some level, i.e., the retention stops increasing with discharge time, or that continue to increase within the discharge time without saturation [1], [2], [3], [4]. In future long-pulse burning plasma experiments, e.g. ITER, the retention due to non-saturating mechanisms is of most concern due to the regulatory issues of in-vessel tritium. To date, most fuel retention studies in tokamaks have been carried out with carbon PFCs and thus give an extensive database for cross machine comparisons of retention mechanisms [5], [6], [7], [8]. Metal PFC retention studies in tokamaks have been done and find large differences in the details of the permanent trapping mechanisms, but this topic is beyond the scope of this work; see Refs. [9], [10].

Fuel retention due to plasma wall interactions in carbon are classified into four characteristic processes; absorption, bulk diffusion, implantation, and co-deposition. Absorption into the carbon porosity saturates quickly every discharge but also outgases after each discharge and is therefore not a concern for fuel retention [4]. Similarly, bulk diffusion of hydrogenic species into graphite is of minimal concern due to very low diffusion coefficients under typical tokamak PFC surface temperatures [11]. Implantation is a saturable process that depends on incident particle energy and flux, as well as surface temperature. Implantation in graphite leads to a deuterium, D, fluence ∼1021 D/m2 within 20 nm of the surface and a deuterium-to-carbon (D/C) concentration ∼0.4 for surface temperatures <500 °C [11]. Co-deposition occurs through the simultaneous deposition/implantation of the D as a carbon layer grows due to net deposition. This process is caused by ions/neutrals eroded from elsewhere in the vessel being net transported to another region. This scrape-off layer (SOL) transport usually leads to carbon layers forming in the divertor region [12]. The D fuel trapped in such layers is not released except at very high surface temperatures (>700 °C) through extraordinary wall conditioning techniques (e.g. thermo-oxidation [13] or disruptive cleaning [14]). This process is of greatest concern for future devices because: (a) the layer can grow continuously throughout the discharge thus indefinitely trapping fuel at a continuous rate; and (b) the large amount of effort involved in removing the trapped fuel from these layers of increasing depth.

In this paper, we report on a recent study to quantify the fuel retention mechanisms using various global particle balance methods in the all graphite DIII-D device. Specifically, the total retention due to absorption and implantation is determined directly via the techniques reported here, and a maximum co-deposition level is inferred from the remaining particles. A combination of unique factors in this study set it apart from the past experiments on this topic. First, a global particle balance calculated continuously through the discharge is directly compared with a particle balance method where the exhausted particles from the pumping system are measured after the discharge and compared with the total amount of particles injected. Secondly, a vessel bake to near 350 °C after an experiment is preformed in an effort to measure the amount of retained fuel released from non-permanently trapping mechanisms (i.e., not co-deposition). This second bake is intended to start the PFCs with a “clean” surface that does not have what is termed “loosely bound” fuel (i.e., retention due to absorption and/or shallow implantation). After such a bake, by careful measurement, one can then determine the semi-permanent trapped particles; what have been termed “tightly bound” which need extremely high temperature (>700 °C) cleaning, or thermo-oxidation to be removed. A vessel bake, at 350 °C, is also performed before each experiment.

In Section 2, a detailed description of the within shot particle balance as well as the between shot particle balance is given. Results of these analysis methods are given in Section 3 for a range of plasma conditions, including a series of discharges that had a vessel bake afterwards to record the amount of fuel released due to the short-term fuel retention. Section 4 discusses these results as they relate to past studies of fuel retention in tokamaks. Finally, conclusions are drawn in Section 5.

Section snippets

Setup and analysis technique

In this context, a global particle balance is simply a balance of the vessel sources and sinks during a discharge. Specifically, it calculates the amount of particles injected into the vessel and compares them to the particles that go: into the core plasma; to the exhaust system of the vessel; and/or into the carbon PFCs. This balance takes the following familiar form [15], [16], [17], [18]:ΓWALL(t)=ΓIN(t)-QPUMP(t)+dNP(t)dt+dN0(t)dt

Here, ΓWALL(t) is a calculation of the particle retention rate

Analysis and results

The purpose of these experiments is to utilize both the dynamic and static particle balance methods in ELMy H-mode discharges to determine the amount of “loosely-bound” recoverable D and co-deposited D in the graphite PFCs of DIII-D. This effort is in contrast to past DIII-D fuel retention experiments which showed the cryopumps “conditioned” the walls to a degree that between shot GDC was not necessary [17], [20], [21]. The current experiments are also an extension of previous fuel retention on

Discussion

The experiments reported here add to the database of similar results from other devices that have addressed particle balance. Namely, at low core density (i.e., low fueling rates) and with strong divertor pumping, the steady-state phase of the discharge can exhibit no net fuel retention within the experimental uncertainty. There are several explanations for such behavior: (a) the implantation has saturated and/or; (b) the co-deposition rates are so low during the stationary phase of the

Conclusions

The finding that most of the fuel retention occurs in the initial phase of the discharge is consistent with most other tokamaks that have performed global particle balances experiments in a pumped divertor configuration [5], [8], [23], [24]. The dominance of external pumping occurs in the operating regime where low external fueling is needed to maintain a steady density with fGW < ∼0.7 and is consistent with these previous studies. Under these conditions, the divertor is most likely in a high

Acknowledgments

This work was supported by the US Department of Energy under DE-AC05-00OR22725, DE-AC52-07NA27344, DE-FC02-04ER54698, and DE-AC04-94AL85000. EAU would like to acknowledge the fruitful discussions and encouragement from M.A. Mahdavi and P.C. Stangeby. He would also like to recognize the initial encouragement and development of this work on DIII-D by the late W.P. West.

References (33)

  • J. Roth

    J. Nucl. Mater.

    (2009)
  • E. Tsitrone

    J. Nucl. Mater.

    (2007)
  • T. Loarer

    J. Nucl. Mater.

    (2009)
  • W. Möller

    J. Nucl. Mater.

    (1989)
  • J.W. Davis et al.

    J. Nucl. Mater.

    (2009)
  • D.G. Whyte et al.

    J. Nucl. Mater.

    (2005)
  • V. Philipps

    J. Nucl. Mater.

    (2009)
  • T. Nakano

    J. Nucl. Mater.

    (2007)
  • G. Haas

    J. Nucl. Mater.

    (1999)
  • P.K. Mioduszewski et al.

    J. Nucl. Mater.

    (2001)
  • D.G. Whyte

    J. Nucl. Mater.

    (1999)
  • B. Lipschultz

    Nucl. Fusion

    (2007)
  • A. Loarte

    Nucl. Fusion

    (2007)
  • T. Loarer

    Nucl. Fusion

    (2007)
  • V. Philipps

    Plasma Phys. Control. Fusion

    (2003)
  • T. Nakano

    Nucl. Fusion

    (2008)
  • View full text