Search

This paper focuses on scrape-off layer and divertor modelling of the medium-density single-null scenario of the Divertor Test Tokamak facility (DTT), under construction in Italy. The modelling was performed using the 2D coupled fluid-Monte Carlo code SOLEDGE2D-EIRENE. For DTT pump designing, neutral pressure at the pump aperture below the dome is calculated in deuterium-only cases as well as with impurity seeding with various puffing levels. This scenario analysis also allowed the characterization of detachment in DTT and the influence of pumping on detachment itself. Two different radiating impurities, neon and nitrogen, were tested in the high power scenario to evaluate the minimum impurity concentration required to achieve sustainable conditions at DTT divertor. The sensitivity of the model was studied by varying the impurity concentration; the model shows a hysteresis-like behaviour between the impurity influx and the total impurity content by which detachment is strongly influenced.

An H-mode database of ASDEX Upgrade plasmas with improved diagnostics for a variety of different plasma conditions has been analysed to study the momentum removal in the Scrape-Off Layer (SOL). A strong reduction, up to a factor of 100, of the electron pressure is observed close to the separatrix. Sets of L-mode and H-mode like plasma simulations with the SOLPS5.0 code package have supported the interpretation of the experimental observations (e.g., apparent momentum gain, strong radial position dependence) and helped to clarify the role of ion-neutral interactions. The experimental data are in turn used to check the SOLPS predictions, which differ quantitatively, suggesting that important pieces of physics are still not successfully captured by the code. Keywords: Momentum removal, Scrape-off layer, SOLPS, Divertor detachment

The impact of intrinsic and actively injected impurities on plasma edge conditions is crucial for optimizing fusion devices like Wendelstein 7-X. The 3D magnetic structure of the scrape-off layer (SOL) plays an essential role in determining the effect of impurities on the plasma edge. It is significant to align active gas injection with the island divertor geometry to improve effectiveness, as injected gases can manipulate neutral fueling behavior and steer heat flux, thereby facilitating improved fueling in fusion devices. Numerical analysis with EMC3-EIRENE and experiments were employed to investigate the impact of radiative edge cooling by impurities on transport in the plasma SOL and at the core-edge interface. Injection of nitrogen and neon in the plasma edge caused a reduction in edge temperature and enhanced radiation, with nitrogen demonstrating a faster response compared to neon. Initial experiments with island divertors highlighted the importance of gas injection location with respect to the island geometry, a finding confirmed by simulations. [F. Effenberg et al 2019 Nucl. Fusion 59 106020DOI 10.1088/1741-4326/ab32c4], 19th Coordinated Working Group Meeting, Berlin 03/12/2019; https://www.ipp.mpg.de/cwgm2019 https://www.ipp.mpg.de/4612574/19thCWGM_Agenda_ver_11_a.pdf

Stellarator plasma edge modelling Introduction (machine, code) Transport model Local magnetic coordinates Wendelstein-7X vacuum field structure Mesh generation procedure Numerical approach Temperatures, density and velocity distributions, {"references":["G. Pełka et al 2019 Problems of Atomic Science and Technology 119(1):27-29","F. Effenberg et al 2017 Nucl. Fusion 57 036021"]}

The first divertor was installed in the JET machine between 1992 and 1994 and was operated with carbon tiles and then beryllium tiles in 1994-5. Post-mortem studies after these first experiments demonstrated that most of the impurities deposited in the divertor originate in the main chamber, and that asymmetric deposition patterns generally favouring the inner divertor region result from drift in the scrape-off layer. A new monolithic divertor structure was installed in 1996 which produced heavy deposition at shadowed areas in the inner divertor corner, which is where the majority of the tritium was trapped by co-deposition during the deuterium-tritium experiment in 1997. Different divertor geometries have been tested since such as the Gas-Box and High-Delta divertors; a principle objective has been to predict plasma behaviour, transport and tritium retention in ITER. Transport modelling experiments were carried out at the end of four campaigns by puffing C-13-labelled methane, and a range of diagnostics such as quartz-microbalance and rotating collectors have been installed to add time resolution to the post-mortem analyses. The study of material migration after D-D and D-T campaigns clearly revealed important consequences of fuel retention in the presence of carbon walls. They gave a strong impulse to make a fundamental change of wall materials. In 2010 the carbon divertor and wall tiles were removed and replaced with tiles with Be or W surfaces for the ITER-Like Wall Project. For complete list of authors see http://dx.doi.org/10.1016/j.fusengdes.2018.10.002

Plasma turbulence intermittency is investigated in the Texas Helimak, a simple magnetic torus which is a physical approximation to the sheared cylindrical slab. The statistical properties of the intermittency are examined through a conditional analysis of probe signal bursts and the existence of so called "blobs", coherent plasma objects which transport large amounts of particles and heat, is explored. It is shown that the conditional analysis most likely extracts artifacts due to the finite ensemble size and that no coherent part of the bursts is detected throughout the Helimak radii for all the externally applied biases. The probe signal bursts rarely follow Poisson statistics, and a match with a recently developed stochastic model based upon this is more than likely due to the large number of parameters.

The edge and scrape-off layer (SOL) plasma of the inter-ELM phase of an H-mode discharge from the TCV tokamak is modeled with the transport code SolEdge2D-EIRENE (Bufferand et al. Nuclear Fusion 55 (2015)). The numerical simulations, in presence and in absence of C impurities sputtered from the first wall, are presented and compared with the experiments, finding an overall good agreement. The application of the standard two-point model to the simulation results leads to an apparent momentum gain along the divertor leg. A two-fluid two-point model featuring thermally decoupled ions and electrons is introduced and applied to the simulation results, overcoming this apparent discrepancy. Keywords: Scrape-off layer, Plasma, Simulations, Transport, Two-point model, TCV

EDGE2D-EIRENE (the ‘code’) simulations show that radial electric field, E r, in the near scrape-off layer (SOL) of tokamaks can have large variations leading to a strong local E × B shear greatly exceeding that in the core region. This was pointed out in simulations of JET plasmas with varying divertor geometry, where the magnetic configuration with larger predicted near SOL E r was found to have lower H-mode power threshold, suggesting that turbulence suppression in the SOL by local E × B shear can be a player in the L–H transition physics (Delabie et al 2015 42nd EPS Conf. on Plasma Physics (Lisbon, Portugal, 22–26 June 2015) paper O3.113 (http://ocs.ciemat.es/EPS2015PAP/pdf/O3.113.pdf), Chankin et al 2017 Nucl. Mater. Energy 12 273). Further code modeling of JET plasmas by changing hydrogen isotopes (H–D–T) showed that the magnitude of the near SOL E r is lower in H cases in which the H-mode threshold power is higher (Chankin et al 2017 Plasma Phys. Control. Fusion 59 045012). From the experiment it is also known that hydrogen plasmas have poorer particle and energy confinement than deuterium plasmas, consistent with the code simulation results showing larger particle diffusion coefficients at the plasma edge, including SOL, in hydrogen plasmas (Maggi et al 2018 Plasma Phys. Control. Fusion 60 014045). All these experimental observations and code results support the hypothesis that the near SOL E × B shear can have an impact on the plasma confinement. The present work analyzes neutral ionization patterns of JET plasmas with different hydrogen isotopes in L-mode cases with fixed input power and gas puffing rate, and its impact on target electron temperature, T e, and SOL E r. The possibility of a self-feeding mechanism for the increase in the SOL E r via the interplay between poloidal E × B drift and target T e is discussed. It is also shown that reducing anomalous turbulent transport coefficients, particle diffusion and electron and ion heat conductivities, leads to higher peak target T e and larger E r, suggesting the possibility of a positive feedback loop, under an implicitly made assumption that the E × B shear in the SOL is capable of suppressing turbulence.

International audience; A control-oriented approach to the monitoring of wall power flux densities on ITER has been successfully developed. It is based on real-time equilibrium reconstruction in 2-D which is then used to describe the deposited heat flux as a poloidal flux function with user specified parameters for the power exhausted into the scrape-off layer (SOL) and the SOL heat flux width. To account for the real 3-D geometry of the plasma-facing components (PFC), appropriate weighting factors are derived from magnetic field line tracing in 3-D. Integration of the 3-D effect is performed with a new GUI-based software environment, SMITER, incorporating a field line tracer and permitting import and meshing of PFC CAD models. The paper discusses the experimental demonstration of the model-based wall heat flux algorithm on the TCV tokamak, reporting on the benchmarking of the new code package, SMITER against infra-red camera heat flux measurements and the derivation of the component shaping weighting factors. A comparison of the real-time estimation of the peak power flux and its spatial location against the off-line infra-red measurement for limiter plasma configurations is presented.

Applying gyrokinetic simulations for theoretical turbulence and transport studies to the plasma edge and scrape-off layer (SOL) presents significant challenges. To in particular account for steep density and temperature gradients in the SOL, the "full-f" code PICLS was developed. PICLS is a gyrokinetic particle-in-cell (PIC) code and is based on an electrostatic model with a linearized field equation and uses kinetic electrons. In previously published results we were applying PICLS to the well-studied 1D parallel transport problem during an edge-localized mode (ELM) in the SOL without collisions. As an extension to this collision-less case and in preparation for 3D simulations, in this work a collisional model will be introduced. The implemented Lenard-Bernstein collision operator and its Langevin discretization will be shown. Conservation properties of the collision operator as well as a comparison of the collisional and non-collisional case will be discussed., Comment: 6 pages, 3 figures, PET-17 conference

Plasma Heat Flux to Solid Structures in Tokamaks Petr Vondráček Energy and particle exhaust is the key issue for future fusion reactors based on magnetic plasma confinement, namely for tokamaks. Good understanding of processes controlling plasma heat fluxes impinging the first wall of tokamaks is of a main concern for fusion research. This doctoral thesis is focused on experimental investigation of plasma heat fluxes in different locations of the COMPASS tokamak. A new infrared thermography system was built for this purpose as a part of this PhD research. There are four main topics studied using the system. First of all, a narrow near-SOL heat flux channel with a very steep radial profile is characterised in HFS limited discharges. It is shown that the near-SOL heat flux decay length is consistent with the drift-based model of the SOL heat transport. Secondly, heat loading of leading edges of misaligned limiters is broadly studied. The optical approximation of the heat flux distribution around a poloidal leading edge is confirmed as a valid approach. Heat loading of a magnetically shadowed side of a toroidal limiter gap is observed experimentally for the first time confirming predictions of particle-in-cell code simulations. Thirdly, localised hot-spots caused by the impact of runaway electrons as well as...

High-power millimeter-waves (mmw) in the electron cyclotron range of frequencies are extensively used in fusion devices. They have become essential for a number of applications ranging from plasma heating and current drive to core confinement preservation. In ITER, 24 MW of power are planned to be injected through the equatorial launcher to perform bulk plasma heating and current drive; and through the upper-launcher, mainly to stabilize neo-classical tearing mode (NTM) and prevent plasma disruption. The narrowness of the power deposition is a key feature of NTM stabilization and thus a key requirement for the upper-launcher beam in ITER. Recent calculations have estimated that, because of the long path of the upper-launcher beam in ITER, electron density fluctuations associated with the turbulence in the edge of the plasma could be responsible for a broadening of the mmw-beam, possibly deteriorating NTM stabilization efficiency. In this thesis we investigate mmw-scattering by plasma turbulence experimentally and numerically. In the first part of this thesis, millimeter-wave scattering by plasma turbulence is investigated in the simple magnetized toroidal configuration, which turbulence is known to exhibit universal properties of the turbulence present at the edge of tokamaks. Scattering experiments run on both the TORPEX device and the Tokamak à Configuration variable (TCV) reveal that field-aligned elongated structures of enhanced electron density (blobs) are responsible for fluctuations of the transmitted mmw-power. Using conditional sampling on Langmuir probe measurements, we show that blobs can either increase the locally measured mmw-power or decrease it, depending on their location. In the case of the TORPEX experiments, experimental 2D-time imaging of the electron density from the HEXTIP-U array of Langmuir probes is used to run full-wave simulations of the mmw-beam propagation. The full-wave simulations are found in agreement with the experiments and reveal that blobs have a defocusing effect on the mmw-beam in their wake, leading to local changes in the mmw-power. In the second part of this thesis, mmw-wave scattering is investigated in L-mode confined TCV plasmas. A transmission diagnostic is installed to measured the mmw-power from the X3 beam reaching the floor of the vessel after propagation in the plasma. A set of wall-embedded Langmuir probes located in the inner-wall of the TCV vessel is used to perform conditional sampling on the detected mmw-power signal and identify the effect of blobs present in the scrape-off layer on the mmw-transmission. Similar results to those found in TORPEX are observed: Blobs located in the upper-part of the scrape-off layer are responsible for changes in the transmitted power which can be either positive or negative depending on the blob location. The effect of a blob on the mmw-transmission is investigated numerically. Conditional sampling is performed on the turbulence data from GBS to reconstruct the time-resolved 2D evolution of the electron density in the scrape-off layer associated with the propagation of a blob in TCV. The time-evolution of the detected mmw-power associated with the blob propagation from the full-wave simulations is successfully compared to the experiment. The numerical simulations show that negative fluctuations of the electron density preceding the blob focus the mmw-beam and the blob defocuses the mmw-beam, resulting in partial variations of the mmw-power. Finally, a combination of simulations run with WKBeam and experimental measurements demonstrate that the mmw-beam is 50% broader in TCV due to the edge turbulence than the predicted mmw-beam after propagation in a non-turbulent plasma.

The search for scenarios in which tolerable power exhaust is combined with good confinement must involve high separatrix densities. We present here the first infrared thermography measurements at ASDEX Upgrade in a high density, high confinement H-mode regime in which no type-I ELMs are present. This regime was formerly called type-II ELM or small ELM regime to distinguish it from type-I ELMs. We report on a broadening of the power fall-off length in this regime of up to a factor of four compared to low density inter type-I ELM conditions. This broadening is correlated to an increased filament detection rate as well as an increase in the pressure close to the separatrix. The broadening of the fall-off length decreases the peak heat flux, while the filaments lead to a quasi-continuous transport into the scrape-off layer and onto the divertor.

Alternating radiation phenomena between the inner and outer divertor regions in the sub kHz range are investigated in the tokamak ASDEX Upgrade. While the inner divertor oscillates between the onset and fluctuating state of detachment, the outer divertor oscillates between conditions where it can maintain high recycling conditions or not. The detachment state of the inner divertor determines the magnitude of the neutral flux through the private flux region, thus sets the recycling conditions at the outer divertor. In return these recycling conditions determine the particle content in the divertor, hence the detachment state at the inner divertor.

Plasma filaments have been observed by reciprocating electric probes in the Scrape-Off Layer (SOL) of the Wendelstein 7-X stellarator. Comparison with target probes indicates that a filament observed in the W7-X SOL extends to the sheath. Two-dimensional simulations of seeded filaments exhibit good quantitative agreement with experimental measurements in filament velocity scalings, despite an assumption of constant field line curvature. Both experiment and simulation show a slow radial propagation of filaments, indicating that filaments are essentially bound to their flux surface and do not perform ballistic radial motion. In contrast, the poloidal propagation along flux surfaces is much faster than the radial motion.

openaire: EC/H2020/633053/EU//EUROfusion We report new results of full-torus gyrokinetic simulations of electrostatic turbulence with Elmfire spanning from the magnetic axis to the scrape-off layer (SOL). The new implementation presented here uses the logical boundary condition, which allows for improved stability and flexibility in terms of geometry. We simulate the full plasma of the FT-2 tokamak (Ioffe Institute, Saint-Petersburg, Russian Federation), with two poloidal limiters defining the SOL. We recover expected results in the SOL and find an improvement in our capacity to model the experimental particle and energy sinks in the SOL.

International audience; Hotspot generation by lower-hybrid (LH) launchers is found to be governed by a resonance in the plasma electric field response to the external drive. The kinetic analysis in 1D-1V in the parallel direction allows one to compute the amplification effect for small amplitude of the external drive. The resonant Lorentzian response distorts the distribution function. An island structure is formed in the suprathermal part at the phase velocity of the external electrostatic drive. The non-linear features enhance the plasma response, driving overlap effects between multiple waves at rather low amplitude. The onset of a plateau in the distribution function, with extent reaching one thermal velocity, is already obtained when the standard overlap condition is achieved. The sensitivity of the resonance to plasma parameters and large variation of the amplification magnitude can compensate the fast radial decay of the small-scale features generated by the LH launchers, which are responsible for the interaction with the cold electrons. This mechanism can trigger hotspot generation further in the scrape-off layer than otherwise expected.

This paper deals with a divertor coil currents optimized procedure to design High Flux Expansion (HFE) configurations in the JET tokamak aimed to study the effects of flux expansion variation on the radiation fraction and radiated power re-distribution. A number of benefits of HFE configuration have been experimentally demonstrated on TCV, EAST, NSTX and DIII-D tokamaks and are under investigation for next generation devices, as DEMO and DTT. The procedure proposed here exploits the linearized relation between the plasma-wall gaps and the Poloidal Field (PF) coil currents. Once the linearized model is provided by means of CREATE-NL code, the divertor coils currents are calculated using a constrained quadratic programming optimization procedure, in order to achieve HFE configuration. Flux expanded configurations have been experimentally realized both in ohmic and heated plasma with and without nitrogen seeding. Preliminary results on the effects of the flux expansion variation on total power radiation increase will be also briefly discussed. © 2018 Elsevier B.V.

Tangentially viewing fast camera footage of the low-field side snowflake minus divertor in TCV is analysed across a four point scan in which the proximity of the two X-points is varied systematically. The motion of structures observed in the post- processed movie shows two distinct regions of the camera frame exhibiting differing patterns. One type of motion in the outer scrape-off layer remains present throughout the scan whilst the other, apparent in the inner scrape-off layer between the two nulls, becomes increasingly significant as the X-points contract towards one another. The spatial structure of the fluctuations in both regions is shown to conform to the equilibrium magnetic field. When the X-point gap is wide the fluctuations measured in the region between the X-points show a similar structure to the fluctuations observed above the null region, remaining coherent for multiple toroidal turns of the magnetic field and indicating a physical connectivity of the fluctuations between the upstream and downstream regions. When the X-point gap is small the fluctuations in the inner scrape-off layer between the nulls are decorrelated from fluctuations upstream, indicating local production of filamentary structures. The motion of filaments in the inter-null region differs, with filaments showing a dominantly poloidal motion along magnetic flux surfaces when the X-point gap is large, compared to a dominantly radial motion across flux-surfaces when the gap is small. This demonstrates an enhancement to cross-field tranport between the nulls of the TCV low-field-side snowflake minus when the gap between the nulls is small., Comment: Accepted for publication in Plasma Physics and Controlled Fusion

The effect of magnetic geometry on scrape-off layer (SOL) transport and detachment behaviour is investigated on the TCV tokamak with the goal of assessing the potential of alternative divertor geometries and for the validation of theoretical models. L-mode experiments reveal that increasing connection length and hence divertor volume by either increasing poloidal flux expansion or divertor leg length have different effects on the boundary plasma. In attached conditions, the SOL heat flux width q inferred from target infrared thermography measurements is weakly dependent on poloidal flux expansion but increases approximately with the square root of the divertor leg length. The divertor spreading factor S shows no clear trend with leg length but decreases with flux expansion. TOKAM3X turbulence simulations of the leg length scan are in qualitative agreement with the experiment and can explain observations by a strongly asymmetric (ballooning) transport at and below the X-point. Evidence for increased transport in the region of low poloidal field is obtained in the Snowflake minus geometry. The presence of an additional X-point in the low-field side SOL increases the effective SOL width by approximately a factor two. Increasing flux expansion and leg length both result in enhanced divertor radiation levels, with the effect being much larger in the latter case. This behaviour, together with the observed trend in q, is consistent with a substantial drop in the density threshold for divertor detachment with increasing leg length and a weak variation with flux expansion. Novel spectroscopic techniques reveal that the drop in target ion current and access to detachment is caused by a reduction of the divertor ionization source due to power starvation, while volume recombination is only a small contributor. This interpretation is confirmed by SOLPS modelling. TCV alternative divertor studies are being extended to neutral beam heated H-mode plasmas. The H-mode power threshold is found to vary weakly between standard, X-, and Super- X geometries. In all cases, ELMy H-mode is obtained at intermediate current, while the discharges are ELM-free at high current. Signs of detachment have so far only been observed in the latter case. Ongoing experiments further investigate H-mode detachment in these plasmas and will be extended to Snowflake configurations.

The turbulent plasma dynamics in the periphery of a fusion device plays a key role in determining its overall performance. In fact, the periphery controls the heat load on the vessel walls, the plasma confinement, the level of impurities in the core, the plasma fuelling and the removal of fusion ashes. Hence, understanding and predicting the plasma turbulence in this region is of crucial importance for the success of the fusion program. The GBS code has been developed in past years to simulate plasma turbulence in the periphery of limited tokamaks. The goal of the present thesis is to extend GBS to the treatment of diverted scenarios. Such configurations are of interest for present state-of-the-art experiments and future fusion reactors. For the implementation of this geometry, we express the model in toroidal coordinates, abandoning the flux coordinates previously used in limited configuration, and overcoming the singularity that this coordinate system presents at the X-point of diverted configurations. The accuracy of the numerical scheme is improved by upgrading the second order finite differences scheme to fourth order on staggered grids. The resulting version of GBS is carefully verified through a series of tests (i.e., a benchmark with the previous version of GBS in limited configuration, a rigorous check of the correctness of the code implementation with the method of manufactured solutions, and a convergence study on a relatively simple diverted configuration). The results of a GBS simulation is then used to investigate the dynamics of coherent turbulent structures, called blobs, that characterise plasma turbulence in the periphery of fusion devices. A diverted double-null configuration is considered, and the blob motion is studied using a pattern recognition algorithm. The velocity of the blobs in the presence of an X-point matches the analytical scaling that we derived by considering the different blob properties in the divertor and main SOL regions, retaining the correction terms that account for blob density and ellipticity. In addition, we show that the blob current pattern observed in the simulation results match the theoretical expectations. Finally, the new version of GBS is run with a realistic diverted magnetic equilibrium, taken from an experiment carried out on the TCV tokamak. First insights of the turbulence properties are in good agreement with the current physical understanding of plasma dynamics in the periphery of diverted tokamaks.

The kinetic code for plasma periphery (KIPP) was used to assess the importance of the kinetic effects of parallel electron transport in the scrape-off layer (SOL) and divertor of JET high radiative H-mode inter-ELM plasma conditions with the ITER-like wall and strong nitrogen (N2) injection. Plasma parameter profiles along a magnetic field from one of the EDGE2D-EIRENE simulation cases were used as an input for KIPP runs. Profiles were maintained by particle and power sources. KIPP generated electron distribution functions, f e, parallel power fluxes, electron-ion thermoforces, Debye sheath potential drops and electron sheath transmission factors at divertor targets. For heat fluxes in the main SOL, KIPP results showed deviations from classical (e.g. Braginskii) fluxes by factors typically of ∼1.5, sometimes up to 2, with the flux limiting for more upstream positions and flux enhancement near entrances to the divertor. In the divertor, at the same time, for radial positions closer to the separatrix, very large heat flux enhancement factors of up to ten or even higher, indicative of a strong nonlocal heat transport, were found at the outer target, with heat power flux density exhibiting bump-on-tail features at high energies. Under such extreme conditions, however, contributions of conductive power fluxes to total power fluxes were strongly reduced, with convective power fluxes becoming comparable, or sometimes exceeding, conductive power fluxes. Electron-ion thermoforce, on the other hand, which is known to be determined mostly by thermal and subthermal electrons, was found to be in good agreement with Braginskii formulas, including the Z eff dependence. Overall, KIPP results indicate, at least for the plasma conditions used in this modelling, a sizable, but not dominant, effect of kinetics on parallel electron transport.

The far scrape-off layer transport is studied in ASDEX Upgrade H-mode discharges with high divertor neutral density N 0 , div , high power across the separatrix P sep and nitrogen seeding to control the divertor temperature. Such conditions are expected for ITER but usually not investigated in terms of turbulent SOL transport. At high N 0 , div and P sep the H-mode discharges enter a regime of high cross-field particle and power transport in the SOL which is accompanied by a significant change of the turbulence characteristic analogous to the transition from conductive to convective transport in L-mode. Parallel particle and power flux densities of several 10 23 m−2 s−1 and 10 MW m−2 have been detected about ∼40 to 45 mm outside the separatrix mapped to the outer mid-plane. The particle flux fall-off length reached up to 45 mm. This paper presents for the first time an empirical condition to enter the high transport regime in H-mode and the relation of this regime to changes in the filamentary transport.

This paper presents two discretisation schemes for the parallel gradient operator used in scrape-off layer plasma turbulence simulations. First, a simple model describing the propagation of electrostatic shear-Alfven waves, and retaining the key elements of the parallel dynamics, is used to test the accuracy of the different schemes against analytical predictions. The most promising scheme is then tested in simulations of limited scrape-off layer turbulence with the flux-driven 3D fluid code GBS (Ricci et al., 2012): the new approach is successfully benchmarked against the original parallel gradient discretisation implemented in GBS. Finally, GBS simulations using a radially varying safety profile, which were inapplicable with the original scheme are carried out for the first time: the well-known stabilisation of resistive ballooning modes at negative magnetic shear is recovered. The main conclusion of this paper is that a simple approach to the parallel gradient, namely centred finite differences in the poloidal and toroidal direction, is able to simulate scrape-off layer turbulence provided that a higher resolution and higher convergence order are used.