Chemical stimulation techniques for geothermal wells: experiments on the three-well EGS system at Soultz-sous-Forêts,France

Rock matrix stimulation is a method of enhancing well production or injection within a broad range of challenging environments, varying from naturally fractured limestones to sandstones with complex mineralogy. A common and often successful stimulation option, matrix acidizing, utilizes acids that react and remove mineral phases restricting fluid flow. Reviewed is the technology of chemical treatments available for oil, gas and geothermal wells and the key elements and results of the chemical reservoir stimulation program at the Soultz-sous-Forêts, France, Enhanced Geothermal System Project.     

Scale formation on polypropylene and copper tubes in mildly supersaturated tap water

This paper presents an experimental study of scaling of polypropylene and copper tubes exposed to flowing tap water over an eight week period. The study was motivated by the recent interest in switching from copper to polymeric materials in heat exchangers and absorbers used in solar water heaters. The chemistry of the tap water is controlled to provide mildly supersaturated conditions at pH equal to 8 and 40 °C. Supersaturation is 5.0 with respect to calcite and 7.9 with respect to hydroxyapatite. These experimental conditions are a departure from the common practice of accelerated testing with distilled water highly saturated with respect to calcium carbonate and thus better represent conditions expected in solar systems. The microstructure and composition of the scale as it grows are examined with scanning electron microscopy, X-ray diffraction, X-ray photoelectron spectroscopy and ion chromatography. The results show scaling in tap water is slower than scaling in distilled or deionized water with a single supersaturated component. In the case of copper, prolonged exposure to tap water oxidizes the surface of the tube and subsequently affects scaling. The oxidized copper surface is more susceptible to calcium phosphate scaling and this tendency results in the higher scaling rate for copper than for polypropylene in water containing phosphate. The scaling rates for polypropylene and copper tubes are 0.8 × 10⁻¹⁰ and 1.7 × 10⁻¹⁰ (mol Ca)/(m²s), respectively.     

Rapid inversion of eddy current data for conductivity and thickness of metal coatings

A feature-based method that determines the thickness and electrical conductivity of a coating on a metal plate from the change in the frequency-dependent impedance of an eddy-current probe coil is presented. Recently a least-squares solution of this problem was presented, which, however, requires approximately 20 CPU minutes on a DEC 5000 work station for the analysis of each set of measurements. We show that a feature-based approach can reduce the time to a few seconds on the same processor. We start by showing that a three-parameter scaling of the resistive component of the impedance change vs. frequency leads to a simple and nearly universal curve. Consequently these parameters provide a simple and compact way of expressing the data. Next, we show that the three scaling parameters can be used to construct a look-up table that determines the conductivity and thickness of the coating. Finally, we test the method using experimental data.     

低渗透油田油层结垢机理：砂岩微观孔隙模型实验研究

目前油层结垢机理研究中普遍采用岩心单相流动实验方法，但实验中不仅无法直接观察岩心结垢过程，而且同时注入两种不配伍盐不也与油田注水过程相差较大。以低渗透安塞油田为例，采用与油田注水过程更为接近的砂岩微观孔隙模型单相驱替实验技术，通过直接观察模型结垢过程并结合渗透度测定，对低渗透油田油层结垢机理进行了研究。     

对高炉煤气洗涤水系统结铅锌铝垢的探讨

针对酒钢高炉煤气洗涤水系统结铅锌铝垢问题，进行了分析探讨，并提出了解决该问题的办法。     

基于DSP的语音信号时域谐波压缩算法

主要讨论了一种将语音信号采样率从标准的8K/s降低到4K/s，同时保证64Kbit/s的数码率和良好的语义可识别性，从而使中低码率语音信号传输得以实现的语音压缩编码方案，实现了该方案涉及到的时域谐波压缩算法，并讨论了如何在DSK Plus上加以实现。     

超声波对硫酸钡结垢影响规律研究

试验结果表明,超声波技术在集输管线和计量站应用中,能够满足在线运行的工况要求,具有很好的防垢与除垢效果。近些年,我国在超声波防垢和除垢技术领域取得显著的成就,该技术具有重要的科研价值和应用前景。其中,超声波的空化效应、活化效应、剪切效应可以有效的除垢防垢。近年来,研究超声波对碳酸钙结垢影响规律的文献比较多,但是超声波对硫酸钡结垢影响规律的研究较少,文章通过称重法和电镜粒度分析法研究超声波对硫酸钡结垢影响规律,为油气田有效的除去硫酸钡污垢提供一定的理论基础。     

多智能体自组织语音识别

语音感知是无人系统的重要组成部分,已有的工作大多集中于单个智能体的语音感知,受噪声、混响等因素的影响,性能存在上限。因此研究多智能体语音感知,通过多智能体自组织、相互协作,提高感知性能非常必要。假设每个智能体输出一个通道的语音流条件下,本文提出一种多智能体自组织语音系统,旨在综合利用所有通道提高感知性能;并进一步以语音识别为例,提出能处理大规模多智能体语音识别的通道选择方法。基于Sparsemax算子的端到端语音识别流注意机制,将带噪通道权重置零,使流注意力具备通道选择能力,但Sparsemax算子会将过多通道权重置零。本文提出Scaling Sparsemax算子,只将带噪较强的通道权重置零;同时提出了多层流注意力结构,有效降低了计算复杂度。在30个智能体的无人系统环境下,基于conformer架构的识别系统实验结果表明,在通道数失配的测试环境下,提出的Scaling Sparsemax在仿真数据集上的文字差错率(WER)相比Softmax降低30%以上,在半真实数据集上降低20%以上。     

Large area CIGS2 thin film solar cells on foils: nucleus of a pilot plant

In earlier research, conversion efficiency of 10.4% (AM1.5) and 9.9% (AM0) has been achieved on small area CuIn_xGa_1−xS₂ (CIGS2) solar cell on 127 μm thick stainless steel substrate. The area of research is mainly focused on studying CIGS2 thin films as solar cell absorber material and growing high efficiency cells on ultralightweight and flexible metallic foils such as 127 μm thick stainless steel and SiO₂ coated 25 μm thick Ti foils. This paper presents the scaling up process of CIGS2 thin film substrate from 2.5 × 2.5 cm² to 10 × 10 cm². Initial scaling up efforts focused on achieving uniform thickness and stress-free films. Process of scaling up consisted of refurbishment of selenization/sulfurization furnace, design and fabrication of scrubber and enlargement of new CdS deposition setup. The scaling up from 2.5 × 2.5 cm² to 10 × 10 cm² substrate size has laid the foundation for PV Materials Lab of Florida Solar Energy Center becoming the nucleus of a pilot plant.