In the frequency range from millihertz to hundreds of megahertz, many different physical and physico‐chemical processes contribute to the electrical polarization of porous water‐bearing rocks. This makes the interpretation of their electrical spectra a complicated problem and requires both elaborate theories and model experiments. At high frequencies, the Maxwell–Wagner–Bruggeman–Hanai (MWBH) theory of effective media, which takes into account only bulk properties, shape and partial volume of components, is very appropriate. At low frequencies, surface films, polarization of the electrical double layer (EDL) and clustering of conductive components can produce very strong polarization; corresponding theoretical models are considered in a companion paper (Chelidze & Gueguen 1999, hereafter referred to as Paper I). This paper is devoted to the review of experimental data and their comparison with theoretical models.  Experiments on saturated mineral powders and rocks with various surface areas and surface chemistries confirm the existence of significant surface contributions to the electrical spectra of conductivity and polarization of water‐bearing rocks and the dominance of this contribution over MWBH values at low frequencies. The effective dielectric constant of porous saturated rocks increases with the surface‐to‐volume ratio of the system and strongly depends on the surface charge (ζ potential). At ζ potential, equal to zero, the low‐frequency dielectric permittivity (DP) is minimal. The experimental data on relaxation times and the magnitude of the surface polarization of water‐bearing porous systems can be satisfactorily explained by theories of film polarization, diffusional polarization generated by deformation of an ‘open’ electrical double layer (EDL) and percolation.

No abstract

Summary A fracture-acidizing treatment of carbonate formations can be considered successful when a relatively good fracture conductivity remains after treatment. To reach such a goal, an uneven etching of the fracture by acid is expected, so that channels are created that maintain the fracture hydraulically open even after "mechanical" closure, and therefore enhance productivity. Residual conductivity is the consequence of uneven etching of the surface, but the way this etching occurs in the field is not well understood and therefore poorly described. We thus propose in this paper an experimental method aiming at defining a methodology to investigate and quantitatively characterize how acid injection conditions affect the fracture surfaces, and how fracture conductivity can be estimated from the so-created surface topography. The statistical investigation of surface topography associated with acidizing experiments proposed in this paper, offers the possibility to evaluate the best fluid formulation and flow rate for a given formation type under well conditions. The field application of this method is evident, since it provides a new and interesting tool for selecting fluid and forecasting the behavior of the fracture after an acid fracturing job. Introduction Acid fracturing is a classical treatment used in carbonate formations to improve well productivity. To reach that aim, acid is injected either at a pressure sufficient to fracture the formation or into an already hydraulically induced fracture. As acid flows along the fracture, it dissolves portions of the fracture faces, generally in a non-uniform manner, so that conductive channels are created that remain "hydraulically" open even after "mechanical" closure. The treatment is thus as successful as the so created fracture is long and conductive. The efficient length of the fracture in a given formation is determined by injection conditions (flow rate), injection fluid composition, acid-formation reactivity (acid spending) and acid fluid loss (or leakoff) from the fracture into the formation. On the other hand, the mechanisms and the conditions that give a conductive acid fracture are poorly evidenced and described in the literature. Acid composition and fluid injection sequences are essential parameters in the design of an acid fracturing treatment. The two jobs described in Table 1, performed on two wells on the ABK field, provide clear evidence of the influence of acid formulation and alternating stages on residual fracture conductivity. Though these treatments were performed in nearly the same formations (dolomites). evaluation of post treatment performance displays (Table 1) a better efficiency of the first job. 61% of the injected acid indeed contributed to the etching, whereas for the second job, the major part of the injected volume was lost in the formation because of leakoff effects and consecutive wormholes formation, leading to a worse residual fracture conductivity. Leakoff by decreasing the acid available for the etching of the fracture faces, reduces the treatment efficiency. The need to control acid fluid loss, and the consecutive formation of channels perpendicular to the main flow (wormholes) led to many studies aiming at identifying the mechanisms of wormhole creation. These studies show that depending on injection rate, three different wormholing mechanisms are identified. Though fluid losses can not be avoided in acid fracturing, these studies enable a better understanding of the way acid is consumed through leakoff depending on the fluid rate, and therefore the way they can be reduced. The third important factor affecting fracture geometry is acid spending during injection. Acid spending is mainly controlled by the acid - rock reactivity, that in turns depends on many factors such as injection conditions, acid concentration, composition, formation composition, temperature, fracture width. On the field, success of an acid fracturing job is evaluated from post treatment performances.

While the d.c. electrical properties of a rock, through the volume conductivity, provide information on pore space structure (Formation factor) and pore fluid, the frequency dependence of the electrical parameters allows the characterization of the pore-matrix interfaces. To illustrate these properties, we present here some results of measurements performed on sedimentary rocks (Vosges and Fontainebleau sandstones). These samples have a classical electrical response, i.e. a 'Cole and Cole' frequency behaviour near a certain relaxation frequency and a low-frequency dispersion. We discuss some of the models that have been proposed in the past to explain this behaviour. It appears that most of them are valid near and above the relaxation frequency, but only Dissado and Hill's model seems to be satisfactory over the whole frequency range. The fractal nature of the interfaces, assumed in two of these models but appropriate only for one of them, could also be a convenient explanation for the frequency dependence. Terra Nova, 3,265-275 265 C. RUFFET €T AL.