博文

流量传感器发展的趋势

已有 8692 次阅读 2012-5-11 16:12 |系统分类:科研笔记| 传感器, 流量

流量测量发展趋势

1．流量测量越来越被人们重视；

2．现在有100多种类型的流量计，遵循30多种物理化学定律定理；

3．最受重视发展的是电磁超声科氏质量流量计；

4．多相流及湿气体流量计也得到重视；

5． MEMS技术在流量计型号中推广；

6．流量计工厂和用户有很大发展；

7．计算流体力学Computational Fluid Dynamics (CFD) 在流量测量中应用渐渐推广，有可能在不久的将来代替流量测量仪表及流量标准装置（或某些功能）；

8．流量信号会出现新的处理方法；

9．流量测量功耗减低；

10．发展现场周期检定；

11．新型流量计标准装置产生。

l 电磁流量计

　　电磁流量计是根据法拉弟电磁感应定律制成的一种测量导电性液体的仪表。

　　电磁流量计有一系列优良特性,可以解决其它流量计不易应用的问题,如脏污流、腐蚀流的测量。

　　70、80年代电磁流量在技术上有重大突破,使它成为应用广泛的一类流量计,在流量仪表中其使用量百分数不断上升。

　　优点:

　　(1)测量通道是段光滑直管,不会阻塞,适用于测量含固体颗粒的液固二相流体,如纸浆、泥浆、污水等;

　　(2)不产生流量检测所造成的压力损失,节能效果好;

　　(3)所测得体积流量实际上不受流体密度、粘度、温度、压力和电导率变化的明显影响;

　　(4)流量范围大,口径范围宽;

　　(5)可应用腐蚀性流体。

　　缺点:

　　(1)不能测量电导率很低的液体,如石油制品;

　　(2)不能测量气体、蒸汽和含有较大气泡的液体;

　　(3)不能用于较高温度。

　　应用概况:

　　电磁流量计应用领域广泛,大口径仪表较多应用于给排水工程;中小口径常用于高要求或难测场合,如钢铁工业高炉风口冷却水控制,造纸工业测量纸浆液和黑液,化学工业的强腐蚀液,有色冶金工业的矿浆;小口径、微小口径常用于医药工业、食品工业、生物化学等有卫生要求的场所。

发展直径小于1mm的；

对流场不敏感的；

能测量非充满管道的；

可测流体导电率小于0.05μs的

目前紧盯的国外厂家是可罗尼。

l 超声流量计

超声流量计是通过检测流体流动对超声束(或超声脉冲)的作用以测量流量的仪表。

　　根据对信号检测的原理超声流量计可分为传播速度差法(直接时差法、时差法、相位差法和频差法)、波束偏移法、多普勒法、互相关法、空间滤法及噪声法等。

　　超声流量计和电磁流量计一样,因仪表流通通道未设置任何阻碍件,均属无阻碍流量计,是适于解决流量测量困难问题的一类流量计,特别在大口径流量测量方面有较突出的优点,近年来它是发展迅速的一类流量计之一。

　　优点:

　　(1)可做非接触式测量;

　　(2)为无流动阻挠测量,无压力损失;

　　(3)可测量非导电性液体,对无阻挠测量的电磁流量计是一种补充。

　　缺点:

　　(1)传播时间法只能用于清洁液体和气体;而多普勒法只能用于测量含有一定量悬浮颗粒和气泡的液体;

　　(2)多普勒法测量精度不高。

　　应用概况:

　　(1)传播时间法应用于清洁、单相液体和气体。典型应用有工厂排放液、:怪液、液化天然气等;

　　(2)气体应用方面在高压天然气领域已有使用良好的经验;

　　(3)多普勒法适用于异相含量不太高的双相流体,例如:未处理污水、工厂排放液、脏流程液;通常不适用于非常清洁的液体。

发展趋势

提高准确度；

降低对流场的敏感；

降低外卡式流量计对管壁的敏感；

扩展温度、压力；

应用神经网络技术；

解决大口径的检定；

发展时差法，重视多普勒。

l 科氏质量流量计

CMF主测量参量是质量流量，第二测量参量是流体密度，还有附加测量流体温度。还可由质量流量和流体密度派生出测量双组分溶液中溶质的浓度。CMF应用最多的是需要考核质量（对应与体积的mass，而非品质）为目标的计量总量或测量/控制流量，具体说有；贸易结算交接计量或企业内部核算计量；批量生产(batch process)进料的分批计量（替代以前费工费时的称重计量）；管道混合(blending)配比的控制。
     然而CMF的零漂等问题限制了一些在贸易计量方向的应用。，列例如美国石油协会（API）在90年代中期还认为CMF在石油工业的运行技术尚不成熟；国际标准化组织石油产品及润滑油委员会石油动态计量分委员会(ISO/TC28/SC2)年会上，因“CMF在石油工业密封管道`输送工艺’中的技术尚不成熟”，撤销专门负责制订CMF国际标准的工作组(WG6)。由于CMF性能进一步完善，在其它领域的贸易交接计量应用方面逐渐增加，现在情况似有变化。
     密度是CMF测量的第二参量，再生产在生产过程中作某些品质指标控制，如溶液稀释程度，交接时防止卖方有意稀释；或求去取溶液中溶质浓度，测量溶液中溶容质流量或总量，如油井口流出油水混合液体中油的产量，还可辨别流动中液体种类，分路发送，如区分管系成品液和清洗液交替流动，分送下游不同管道。
    90年代中期CMF又拓展到测量液体的粘度，利用CMF的压力将降与粘度的函数关系辅以差压变送器作在线测量。
    CMF对被测液体的粘度适应范围宽，从低粘度液化石油气到高粘度原油和沥青液。具据国外某仪表厂90年代出初统计分析表明，销售使用于中高粘度液体占50%以上，其中400mPa•s以上占10%。CMF还可应用于非牛顿流体和液固双相流体的流量测量，如乳胶、悬浮高龄高岭土液、巧克力、肉糜浆等。
     早期CMF仅用于液体，然后扩大应用与于高压气体，到90年代初才有适用于测量中低压气体的仪表。据Micro Motion公司称；迄1997年该公司已有7500台CMF应用于气体，其中服务于汽车压缩天然气(CNCCNG)加气站计量的CMF有6000台。
     用户产业分布；：据国外某仪表制造厂90年代初统计分析，CMF的应用中化学工业占40%，石油工业（包括炼制和储运）占20%，食品工业23%，其它占17%，其中食品工业占有相当比列比例；在国内当前石油、石化业用户资金雄厚，用的较多，而食品工业用户可谓绝无仅有。

大部分制造厂以“量程误差加零点不稳定性”的方式表达基本误差。这是一种巧妙的表达方式，给用户产生一种精确度很高的印象。实际上在低流量或接近下限流量时，误差较大，基本误差常超过量程误差一倍以上，选用时应予注意。基本误差通常在±（0.15～0.5）%R之间，重复性误差一般为基本误差的1/4～2/3；流量范围度大部分在（10：1）～（50：1）之间，有些则高达（100：1）～（150：1）。基本误差与范围度有关，列如例如Micro Motion公司D系列10：1时为±0.36%R，20：1时为±0.58%R。
零点不稳定性通常以%FS表示，也有以流量值kg/min表示。零点不稳定性一般再在±（0.01～0.04）%FS之间。若±0.04%FS零点不稳定性和20：1范围度的仪表，下限流量时因零点不稳定误差可能达到该测量点的±0.8%R。
由于CMF精度不断提高，对于精度较低仪表予以忽略的介质温度和静压变化影响将凸显出来。实际工作条件下测量精确度要考虑介质温度附加误差δ_T和静压附加误差δ_P，评估测量条件下测量误差δ_Σ通常由基本误差δ_B和δ_T、δ_P按式3合成。

下文将进一步讨论介质温度和静压的影响。

提到CMF流量范围度很大，实际上是由于上限流量定的得很高所致，与其他类型仪表如容积式、涡轮式相比，如以水的密度计算名义口径流速高达8～12m/s，有些型号甚至达15～16m/s，而容积式和涡轮式仅为3～5m/s。测量管内流速还要高，因此大部分型号CMF的压力损失较大，用于水等低粘度液体时为0.1～0.2MPa，选用时应予注意。
    按使用条件选择CMF规格大小时考虑的主要因素之一为估算仪表压力损失（或称压力降）△pp是否在管系允许值之内。，在允许压力降情况下，为获得最佳测量精确度使用的满度流量尽可能在CMF的流量范围内选的得高些。通常CMF的名义口径小于（或等于）管径，很少大于管径者。
    但也有少数型号仪表压力降较低，列如例如RHM系列上限流量名义口径流速仅2～3m/s，压力降约在0.05MPa左右。

    CMF的压力降随着流体粘度增加而增加。图4所示是D150型（口径40/50mm）的不同粘度流体流量-压力降关系线列图μ=1mPa·s，相当与常温下水粘度，μ=0.01mPa·s相当与于大部分气体的粘度。从图上可以看出粘度为500mPa·s液体的压力降为水的10倍。高粘度液体在仪表中呈层流流动，压力降△p和流量qm之间呈线线性关系（即，式中k为系数，指数n=1），低粘度时为紊流流动,`基本上为平方关系（即n=2），中等粘度关系线为折线，小流量段呈层流，中高流量段为从层流转向紊流的过渡区流动，n在1与2之间。
    所使用流体的粘度在图示线列之间，有文献建议可采用比列比例内插法进近似计算，只适用与于高粘度液体层流流动区。对于中低粘度，粘度-压力损失呈指数关系的紊流区和过渡区并不适用，只能是粗略估计。
    对于在原有管线上以CMF替代其他类型流量仪表（如涡轮流量计）的技术改造，更要核算动力泵扬程能否满足克服CMF所增加压力损失，必要时换较大扬程的泵。

测量气体流量

气体流量的能否测量取决于是否达到规定的质量流量值，由于气体的密度低，必须要在很高的压力和很高的流速下才能达到。列如例如，Micro Motion的DS-100型（25mm口径）仪表流量达到额定流量范围最大值455kg/min时，空气密度若为100kg/m³，绝对压力必须达到7.6MPa，气其流速要高达154.5m/s，即使流量在额定最小流量68kg/min时，流速也需要达23m/s。
     有些型号仪表则规定气体密度下限，列如例如Heinnchs公司的TH系列为2kg/m³；‘Krohne；公司的MFS-3000系列1.5E和10E型（口径分别为6和8mm）为50kg/m³，Foxboro公司，的CFS10=1/AS 系列为200kg/m³，如测量空气流量，则绝对压力必须分别达到4.2和1608MPa16.8MPa。
    同一仪表用于测量气体时性能低与于测量液体。列如例如；制造厂声称EIite系列时测量气体时误差为±0.5%R而测量液体时误差为（±0.1%+零点不稳定度）。但该厂另一论文对同一系列仪表试验结论又称所有数据均优于±2%Rr，读该文所附各图，可见到在测量低压气体时测量误差有接近或超过1%者，这是因为流量处于额定流量百分之几的低流量，是零点不稳定度所起主要作用，低压时重复性也较差，达0.6%。
     用于测量低压气体应注意到可测量流量将大为降低，列如例如EIite系列CMF100型（口径25/40mm）再在测0.175MPa压缩空气时最大流量仅为约4%额定流量。
     通常用于气体的CMF不用气体效验校验，仍用水校效验的仪表的常数，。一般认为两者之间差别不大，实际上还是有些差别的。文献{9}在试验后认为流体密度从1000kg/m3（水）到2kg/m3（0.17MPa空气）很宽的范围内，用工厂校准的仪表常数，精确度优于2%，一般误差小于±0.5%。英国工程实验室（NEL）也曾对6台口径25mm CNF作过液气对比实验；3台在较低压力1.5MPa空气实验，其中2台非线形比液体时大0.6%，1台重复性1%，1台重复性低劣达15%，仪表常数变化10%；3台在较高压力6MPa实验，其中2台重复性比液体时大0.3%，非线形分别比液体大1.5%和1.3%，1台不能工作。

含有气体的液体

制造厂通常声称含有百分之几体积比游离气体的液体带给测量值的影响不大。当测量气泡小而分布均匀的液体，列如例如冰淇淋和相似乳化液，可能是对的。，然而意大利计量院对7种型号CMF含气量影响实验表明；含气泡1%（体积比）时有些型号无明显影响，有些型号误差为1%～2%，而其中某一双管直管式型号则高达10%～15%；含气泡10%时，误差普遍增加到15%～20%，个别型号高达80%。Danfoss公司的实验也证明。当液体含0.3%气泡时，仪表仍可保持原由精确度；当含气量达5%时（在常压下），仪表误差以达10%此外，流体的压力、流速、粘度和气液混合方式等不同带来的影响也不一样。但有些型号CMF声称可测量含气量较高的液体，列如Krohne公司MFS200型（图3（h））所示双并联测量管口径15～25mm仪表，在合适应用条件下含气量可达15%，因此在制造厂未专门说明可测量含有气体的液体时，最好测量可能含有气体液体的仪表前采取脱气措施。

含有固体的液体

测量含有少量固体的液体流量时，各种类型CMF都有较高的信赖度。当固体含量增加，固体具有强磨蚀性或者软固体（如食品汤汁中的蔬菜块），就应按流体的特点选用合适类型测量管的CMF。
含有固体较多或含有软固体，应避免选用测量管内径比名义管径小得多的仪表，防止堵塞。最好选用单管型或双管型中的串联型，因为如用双管型中的并联型，分流器上粘附杂物导致改变二路分流量，产生误差；更为严重的情形是如一路堵塞可能不被立即发现。
测量强磨蚀性的浆液时间同样有堵塞问题，且对分流管的磨蚀不均匀亦会改变原来得的分流比，因此亦不宜选用双管并联型。，最好采用单直管形状测量管管壁较厚的CMF。因为测量管形状复杂易产生管壁磨蚀不均匀。

流体工况或物性参量对流量测量的影响

通常仪表制造厂的样本和使用说明书等技术文件声称CMF的测量性能不受流体的温度、静压、密度、粘度变化影响。，然而随着用户日益增加应用经验，感到并非完全如此，集资委托第三方研究开展影响量的实验研究。制造厂也开展各项应用技术研究，有些影响量已达到可提出修正量的程度。
      如果CMF的流量测量精确度仍为初期的0.5%～1%R，常用范围的流体工况和物性变化影响或可可忽略不计，然而当前精确度指标定的很高，达0.1%～0.15%R，影响量就更凸显出来了。
   温度影响
     介质温度或环境温度变化会改变测量振动管的扬杨氏模量和影响零漂的结构等各种因素。大部分型号仪表对杨扬氏模量的温度系数经电子线路补偿以减少其影响量；零漂影响由于是受振动管几何形状和结构件的非对称性所形成，是不能再现的，因此尚难减小消除。然而杨扬氏模量的温度系数是一个统计量，因测量管材料批号和热处理等工艺的不一致性，有一定幅度范围。有可能补偿不足或过渡，不可能全部补偿为零。各制造厂所提供流体工作温度范围仅根据仪表材料结构等因素来确定的，并不意味着再次在此范围内保持常温下校准的性能。仅有少数制造厂能提供仪表的温度影响量，如Micro Motion公司。
      英国NEL曾对90年代初国外市场上多家制造厂CMF的温度影响量做过实流试验。水温变化范围5～400C。每改变一次水温，仪表在流量试验前调零，在该温度内以后就不能再调。8台仪表中3台无影响，1台仪表常数变化0.5%，2台变化1%～1.5%，2台变化1.5%～2%。5台变化仪表的温度影响量为±（0.014～0.057）%/0C，还是相当大的。
       压力影响
      液体静压增大会使测量振动管呈绷紧（stiffening）现象，弯曲管还有布登管效应（Bour-don effecf），产生一个负向偏差。这两种压力效应虽然影响量很小，但是使用时静压与校准时相差甚大时，对于高精仪表其值还是不容忽视的。压力影响量取决于测量管管径、壁厚和形状，小口径仪表由于壁厚管径比大，影响量小；大口径仪表则壁厚管径比小，影响量大。Micro Motion公司提供该公司仪表压力影响数据，以校准时压力0.2MPa为基准，CMF100型仪表（口径25mm）压力影响量为-0.03%R/MPa，CMF200型（口径40/50mm）为-0.12%R/MPa；D系列较大，D300型（口径80/100mm）为-1.35%MPa，D600（口径150/200mm）为-0.75%MPa。若使用过程中压力有很大变动，则可以根据实际静压修正仪表常数。
     NEL对90年代初市场上8台CMF所作静压影响试验结果如表1所示，静压影响量最大为-1.75%/MPa，最小为-1%/MPa，平均为-1.4%/MPa.

表1 压力影响量

静压 / MPa		2	2.4	2.8
流量测量误差/%	平均	-2.21	-3.25	-3.75
	最小	-1.57	-2.55	-2.6
	最大	-3.15	-4.00	-4.56

注:以校准时压力为基准

    密度影响
    以前认为CMF的流量测量性能不受介质密度影响，但近年各方实验说明还是有一定影响，认为误差小于±0.5%R，其中有密度影响部分。
     NEL 对90年代初市场上8台CMF，以4种不同密度的液体做密度影响试验，密度范围从煤油0.78到乙二醇1.11kg/L。8台仪表中有1台变化+0.5%（以煤油为基准）。
      Danfoss公司对本公司cmf试验也证明存在密度影响。10mm口径仪表介质密度2kg/L的流量值与1kg/L相比，相差-0.1%；0.5kg/L的介质与1kg/L的介质相比为+0.06%。
     粘度影响
     粘度较高的液体会较多吸收科里奥利激励系统的能量，在流动开始时尤甚。这一现象对有些结构设计的CMF可能导致测量暂时停止振动，直到正常流动。
     NEL 对90年代初市场上8台CMF，用水、煤油、粗柴油、乙二醇四种粘度液体，粘度范围为1～29.5mm²/s，在称重流量标准装置上试验。其中一台有明显粘度影响，大流量(2～5kg/s)时仪表常数变化0.25%，在小流量时20%Q_max(0.5kg/s)时变化0.5%，10%时变化2.2%。

安装使用

流量传感器安装一般要求

由于测量管形状及结构设计的差异，同一口径相近流量范围不同型号传感器的重量和尺寸差别很大，列如例如80mm口径者仅45kg，重者达150～200kg。安装要求亦千差万别，因此必须按照制造厂规定的安装方法和趋避禁止事项，列如例如有些型号流量传感器直接连接到管道上即可，有些型号却要求设置支撑架或基础。为隔离管道振动影响仪表，有时后候传感器与管道之间要介以柔性管连接，而柔性管与传感器之间又要一段有支撑件分别固定的刚性直管。选购之前应向拟购CMF的厂商索取安装使用说明书参照比较和选择。
安装设计时尽可能使其有长的使用寿命，为除去过早磨损和产生测量误差的固形物和夹杂气体，按流体和管道条件在传感器上游装过滤器或气体分离等保护装置。，若希望能在现场在线校准仪表，应考虑引流连接口和阀，以及相应的空间。

流量传感器安装姿势和位置

流量传感器测量管内残留固形物、结垢、潴留气体等均将影响测量精度。一般说装于自下面下而上流动的垂直管道较为理想；但对于非直形测量管CMF装在垂直管道还是水平管上。取决于管道振动状况和应用条件。
安装位置必须使测量管内充满液体，列如例如水平管道上流体流过CMF后直接放如入容器而无背压，测量管往往不能充满，会使输出信号激烈波动。

截止阀和控制阀的安装

为使调零时没有流动，CMF上下游设置截止阀，并保证无泄漏。控制阀应装在CMF下游，CMF保持尽可能高的静压，以防止发生气蚀和闪蒸（fIashing）。

脉动和振动

为勿使流程中发生的和外部的机械振动影响CMF，向制造厂询问所提供CMF的共振频率范围，以判断现场脉动或振动频率是否接近CMF的共振频率。亦可向制造厂提供现场振动状况咨询是否需要采取下列措施，如:1）设置脉动衰减器，2）设置振动衰减器或柔性连接管，3）特殊的流量传感器的夹装固定设备，等等。

防止CMF间相互影响

同一型号两台CMF串联安装，或多台CMF接近地并行（或并联）安装，尤其装在同一支撑台架时，测量管振动会使各CMF间相互影响，产生干扰而引起异常振动，严重时使仪表无法工作。安装时应采取防范措施，如；向制造厂提出错开接近仪表的共振频率值；拉开流量传感器距离，不设置在同一台架上，独立设置支撑架；流量传感器异方向安装；流量传感器间设置防振材料隔离等方法。

管道应力和扭曲

CMF 法兰与管道法兰连接旋紧螺栓时要均匀，勿使CMF产生应力（列如例如管道两法兰平面不平行所致）。若在布设管道时预接入与CMF同样长度的短管，可防止不良布管形成的应力。在使用过程中由于工艺流程压力和温度变化，CMF会受到管线轴向力或弯曲/扭曲力。影响测量性能，要做好必要的固定支架。

强磨蚀性浆液的使用

前文提到测量强磨蚀性浆液最好选用直管单管型并且要使测量管处于垂直位置，以免管壁磨损不匀，缩短使用寿命。然而管壁厚度变薄会降低测量管钢刚性而改变流量测量值，因此在这种场所的运行初期要定期检测，确认使用周期。
测量管内壁结构结垢或漂移沉积也会影响测量精确度，因此要定期清洗。

零点漂移和调零

零点漂移来自流量传感器部分，主要原因有；1）机械振动的非对称性和衰减；2）流体的密度粘度变化，影响前者的因素有；a) 管端固定应力的影响； b)振动管钢刚度的变化；c)双管谐振频率不一致性；d)管壁材料的内衰减。后者影响零位的原因是结构不平蘅，因此即使在空管时将双管的谐振频率调整一致，到充满液体时可能产生零漂，同样因粘度引起的振动衰减与频率有关，在流动时亦可能产生零漂。
最后调零必须在安装现场进行，流量传感器排尽气体，充满待测流体后在再关闭传感器上下游阀门，在接近工作温度的条件下调零。安装方面变动或温度大幅度变化时需要重新调整。

　　我国CMF的应用起步较晚,近年已有几家制造厂，自行开发供应市场;现在已有多家制造厂。

开发大口径；

开发直管式；

增加抗振力；

增大温度、压力。

l 热式质量流量计

增加抗附着影响；

提高准确度。

l mems流量计

新技术，发展方向需探索。

l 相关流量计

难度大，但有发展前途。

l 差压式流量计

　差压式流量计是根据安装于管道中流量检测件产生的差压,已知的流体条件和检测件与管道的几何尺寸来计算流量的仪表。

　　差压式流量计由一次装置(检测件)和二次装置(差压转换和流量显示仪表)组成。通常以检测件形式对差压式流量计分类,如孔板流量计、文丘里流量计、均速管流量计等。

　　二次装置为各种机械、电子、机电一体式差压计,差压变送器及流量显示仪表。它已发展为三化(系列化、通用化及标准化)程度很高的、种类规格庞杂的一大类仪表,它既可测量流量参数,也可测量其它参数(如压力、物位、密度等)。

　　差压式流量计的检测件按其作用原理可分为:节流装置、水力阻力式、离心式、动压头式、动压头增益式及射流式几大类。

　　检测件又可按其标准化程度分为二大类:标准的和非标准的。

　　所谓标准检测件是只要按照标准文件设计、制造、安装和使用,无须经实流标定即可确定其流量值和估算测量误差。

　　非标准检测件是成熟程度较差的,尚未列入国际标准中的检测件。

　　差压式流量计是一类应用最广泛的流量计,在各类流量仪表中其使用量占居首位。近年来,由于各种新型流量计的问世,它的使用量百分数逐渐下降,但目前仍是最重要的一类流量计。

　　优点:

　　(1)应用最多的孔板式流量计结构牢固,性能稳定可靠,使用寿命长;

　　(2)应用范围广泛,至今尚无任何一类流量计可与之相比拟;

　　(3)检测件与变送器、显示仪表分别由不同厂家生产,便于规模经济生产。

　　缺点:

　　(1)测量精度普遍偏低;

　　(2)范围度窄,一般仅3:1~4:1;

　　(3)现场安装条件要求高;

　　(4)压损大(指孔板、喷嘴等)。

　　应用概况:

　　差压式流量计应用范围特别广泛,在封闭管道的流量测量中各种对象都有应用,如流体方面:单相、混相、洁净、脏污、粘性流等;工作状态方面:常压、高压、真空、常温、高温、低温等;管径方面:从几mm到几m;流动条件方面:亚音速、音速、脉动流等。它在各工业部门的用量约占流量计全部用量的1/4~1/3。

趋势

发挥优点，智能化；

减小直管段；

减小对流场的依赖；

增大范围度。

l 渦街流量计

涡街流量计(USF)

　　USF在60年代后期进入工业应用,80年代后期起在各国流量仪表销售金额中已占4%~6%。1992年世界范围估计销售量为3.54.8万台,同期国内产品估计在8000~9000台。

　　由上述可知,流量计发展到今天虽然已日趋成熟,但其种类仍然极其繁多,至今尚无一种对于任何场合都适用的流量计。

　　每种流量计都有其适用范围,也都有局限性。这就要求我们:

　　在选择仪表时,一定要熟悉仪表和被测对象两方面的情况,并要兼顾考虑其它因素,这样测量才会准确;

l 浮子流量计

l 涡轮流量计

　　涡轮流量计,是速度式流量计中的主要种类,它采用多叶片的转子(涡轮)感受流体平均流速,从而且推导出流量或总量的仪表。

　　一般它由传感器和显示仪两部分组成,也可做成整体式。

　　涡轮流量计和容积式流量计、科里奥利质量流量计称为流量计中三类重复性、精度最佳的产品,作为十大类型流量计之一,其产品已发展为多品种、多系列批量生产的规模。

　　优点:

　　　(1)高精度,在所有流量计中,属于最精确的流量计;

　　　(2)重复性好;

　　　(3)元零点漂移,抗干扰能力好;

　　　(4)范围度宽;

　　　(5)结构紧凑。

　　缺点:

　　　(1)不能长期保持校准特性;

　　　(2)流体物性对流量特性有较大影响。

　　应用概况:

　　涡轮流量计在以下一些测量对象获得广泛应用:石油、有机液体、无机液、液化气、天然气和低温流体统在欧洲和美国,涡轮流量计在用量上是仅次于孔板流量计的天然计量仪表,仅荷兰在天然气管线上就采用了2600多台各种尺寸,压力从0.8~6.5MPa的气体涡轮流量计,它们已成为优良的天然气计量仪表。

l 均速管

l V锥流量计

l 层流流量计

l 多相流流量计

参考文献（References）

1. Ashauer, M., Glosch, H., Hedrich, F., Hey, N., Sandmaier, H. and Lang, W. (1999) Thermal flow sensor for liquids and gases based on combinations of two principles, Sensors and Actuators, 73:7-13.

2. Atkinson, I., Berard, M., Hansen, B.-V. and Segeral, G. (1999) New generation multiphase flow meters from Schlumberger and Framo Engineering AS. 17^thNorth Sea Flow Measurement Workshop, Paper 12 pp 154-165.

3. Baker, R. C. (1988/9) Introductory guide to flow measurement, Professional Engineering Press, London.

4. Baker, R. C. (2000) Flow Measurement Handbook, Cambridge University Press, New York.

5. Baker, R. C. (2002) Recent and likely developments in flow measurement, Report CUED/E – MANUF/TR 19, Institute for Manufacturing, Cambridge University, England.

6. Bates, C. J. (1999) Upstream installation and misalignment effects on the performance of a modified electromagnetic flowmeter. J. Flow Meas.Instrum., 10: 79-89.

7. Bates, C. J. (2000) Performance of two electromagnetic flowmeters mounted downstream of a 90 mitre bend/reducer combination. Measurement: Journal of the International Measurement Confederation, 27: 197-206.

Measurement Confederation, 27: 197-206.

8. Battye, J. S. (2001) Acoustic considerations effecting the design of demodulators for the ultrasonic correlation flow meter. Sensors and Actuators, A88(1):29-40.

9. Belhadi, A., Cheesewright, R. and Clark, C.. (2000) The simulation of Coriolis meter response to pulsating flow using a general purpose F.E. code. J. Fluids and Structures 14: 613-634.

10. Bignell, N. and Takamoto, M. (2000) Editorial of a special issue on Sonic Nozzles. J. Flow Meas. Instrum., 11(4): 255.

11. Caetano, E., Pinheiro, J. A., da Costa e Silva, C. B., Kuchpil, C. and Dykesteen, E. (2000) Operational experience with subsea multiphase flow meter. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

12.. Cellos, H. (1999) Multiphase flow measurement system of high-GOR applications. 17th North Sea Flow Measurement Workshop, Paper 1 pp 15-23.

13. Cheesewright, R. and Clark, C. (1998) The effect of flow pulsations on Coriolis mass flow meters. J. Fluids and Structures,12:1025-1039.

14. Cheesewright, R., Atkinson, K. N., Clark, C., ter Horst, G. J. P., Mottram, R. C. and Viljeer, J. (1996) Field tests of correction procedures for turbine flowmeters in pulsatile flows. Flow Meas. Instrum., 7(1): 7-17.

15. Cheesewright, R., Clark, C. and Bisset, D. (2000) The identification of external factors which influence the calibration of Coriolis massflow meters. Flow Meas. Instrum., 11: 1-10.

16. Clarke, D. W. (1998) Non-linear control of the oscillation amplitude of a Coriolis mass-flowmeter. European Journal of Control, 4:196-207.

17. Couput, J. P., de Laharpe, V., Gajan, P. and Strezelecki, A. (1999) Behaviour of venturi meter in two-phase flows. 17th North Sea Flow Measurement Workshop, Paper 6 pp 71-84.

18. Couput, J.-P., Gajan, P., de Larharpe, V. and Strzelecki, A. (2000) Wet gas metering in the upstream area: needs, applications and developments. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

19. Cunningham, T. J. and Hensley, D. P. (2001) Using IMAT and Matlab for Coriolis flowmeter design. Proceedings of the International Modal Analysis Conference (IMAC), Kissimmee, USA, 1: 165-71.16

20. Dahlstr, M. J. (2000) Effortless oil ultrasonic fiscal meter operation: Krohne Altosonic-V, with master meter approach. Vignis-Snorre Crossover and the Snorre B Export Station. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

21. Dane, H. J. and Wilsack, R. (1999) Upstream pipe wall roughness influence on ultrasonic flow measurement. 17th North Sea Flow Measurement Workshop, Paper 9 pp 106-116.

22. Daniel, P., Tudge, M. and Lund, J. (2000) A venturi based wet gas meter with on-line gas mass fraction estimation. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

23. de Carvalho, J. G. and Antunes, B. de C. (2000) Multi-phase flowmetering using twin helicoidal rotors ?the 'multi-stream' meter. FLOMEKO'2000 the 10th International Conference on Flow Measurement, Salvador, Brazil: Paper B4.

24. Donoghue, M. and Crane, M. (1999) Square root error and impulse line pulsation at CATS Terminal, Middlesbrough, UK. 17th North Sea Flow Measurement Workshop, Paper 24 pp 315-345.

25. Dyakowski, T. (1996) Process tomography applied to multi-phase flow measurement. Meas. Sci. Technol., 7: 343-353.

26. Dykesteen, E. (2000) Multiphase meters offshore Malaysia. Journal of Offshore Technology

27. Enoksson, P., Stemme, G. and Stemme, E. (1997) A silicon resonant sensor structure for Coriolis mass-flow measurements. J. Microeletromechanical Systems, 6(2): 119-125.

28. Ginesi, D. (1997) Flow sensing: the next generation, Control Engineering, November: 56-64

29. Gopal, M. and Jepson, P. (1996) Development of a novel non-intrusive, ultrasonic flowmeter for wet gas pipelines. Proceedings of the ASME Fluids Engineering Division Summer Meeting, San Diego, CA, USA, 236(1): 647-652.

30. Hall, A. R. W. and Reader-Harris, M. J. (1999) Use of venturi meters in multiphase flow measurement. 17th North Sea Flow Measurement Workshop, Paper 7 pp 56-105.

31. Hall, A. R. W., Whitaker, T. S. and Millington, B. C. (1997) Multiphase flowmetering: current status and future developments. 29th Offshore Technology Conference, Houston, TX, USA, 4:545-552.

32. Harrison, P. S., Shires, G. L. and Parry, S. J. (1999) The effects of salinity variation on dual energy multiphase flow measurements and Mixmeter homogeniser performance in high gas and high viscosity operation. 17th North Sea Flow Measurement Workshop, Paper 15 pp 193-211.

33. Hartley, P. E., Roach, G. J., Stewart, D., Watt, J. S., Zastawny, H. W. and Ellis, W. K. (1995) Trial of a gamma-ray multiphase flowmeter on the West Kingfish oil platform. Nucl. Geophys. 9(6): 533-552.

34. Henry, M. P., Clarke, D. W., Archer, N., Bowles, J., Leahy, M. J., Liu, R. P., Vignos, J. and Zhou, F. B. (2000) A self-validating digital Coriolis mass-flowmeter: an overview. Control Engineering Practice, 8: 487-506.

35. Hilgenstock, A. and Ernst, R. (1996) Analysis of installation effects by means of computational fluid dynamics - CFD vs experiments? J. Flow Meas. Instrum., 7(3/4):161-171.

36. Holm, M., Stang, J. and Delsing, J. (1995) Simulation of flowmeter calibration factors for various installation effects. Measurement: J. Int. Meas. Confederation,15(4): 235-244.

37. Hulbert, G. M., Darnell, I. and Brereton, G. J. (1995) Numerical and experimental analysis of Coriolis mass flowmeters. AIAA/ASME/ASCE/AHS Structures, Structural Dynamics and Materials Conference ?Collection of Technical Papers, 3: 1889-1893.

38. Karnik, U. (1995) A compact orifice meter/flow conditioner package, Proceedings of the American Gas Association Operating Section, Las Vegas, 95-OP-009:564-585

39. Keita, N. M. (2000) Ab initio simulation of Coriolis mass flowmeter. FLOMEKO'2000 the 10th International Conference on Flow Measurement, Salvador, Brazil: Paper C9.

40. Kutin, J; and Bajsic I. (2001) Stability boundary effect in Coriolis meters. J. Flow Measurement and Instrumentation., 12(1):65-73.

41. Lansing, J. (2000) Smart monitoring and diagnosis for ultrasonic gas meters. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

42. Larson, T. A. (2000) Operational experience and utilisation of the data from subsea multiphase flowmeter in the West Brae Field. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

43. Lawrence, P. A. (2000) Wellhead metering using V-cone technology. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

44. Le Brusquet, L. and Oksman, J. (1999) Improving the accuracy of tracer flow-measurement techniques by using an inverse-problem approach. Meas. Sci. Technol., 10:559-563.

45. Leggett, R. B., Borling, D. C., Powers, B. S., Shehata, K., Halvorsen, M. and AboElenain, A. (1996) Multiphase flowmeter successfully measures three-phase flow at extremely high gas column fractions ?Gulf of Suez, Egypt, European Petroleum Conference, Milan, Society of Petroleum Engineers, SPE 36837:215-226.

46. Letton, W., Svaeren, J. A. and Conort, G. (1997) Topside and subsea experiences with the multiphase flowmeter, Proceedings SPE Annual Technical Conference, 345-357.

47. Liu, Y., Lynnworth, L. C. and Zimmerman, M. A. (1998) Buffer waveguides for flow measurement in hot fluids. Ultrasonics, 36:305-315.

48. Lunde, P., Fr鴜sa, K.-E. and Vestrheim, M. (2000) Challenges for improved accuracy and traceability in ultrasonic fiscal flow metering. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

49. Lynnworth, L. C., Nguyen, T. H., Smart, C. D. and Khrakovsky, O. A. (1997) Acoustically isolated paired air transducers for 50-, 100-, 200-, or 500-kHz applications, IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, 44:1087-1100.

50. McBrien, R. K. (1997) High pressure pulsation effects on orifice meters. ASME Fluids Engineering Division Summer Meeting FEDSM'97, Paper 3700.

51. Menendez, A., Biscarri, F. and Gomez, A. (1998) Balance equations estimation with bad measurements detection in a water supply net. J. Flow Meas.Instrum., 9: 193 ?198.

52. Mi, Y., Ishii, M. and Tsoukalas, L. H. (2001) Investigation of vertical slug flow with advanced two-phase flow instrumentation. Nuclear Engineering and Design, 204(1-3): 69-85.

53. Mickan, B., Wendt, G., Kramer, R. and Dopheide, D. (1996) Systematic investigation of flow profiles in pipes and their effects on gas meter behaviour. Measurement, 22:1-14.

54. Mohamed, P. G. and Al-Saif, K. H. (1998) Field trial of a multiphase flowmeter, Society of Petroleum Engineers Annual Technical Conference and Exhibition, New Orleans, SPE 49161 (also in synopsis in 1998 Journal of Petroleum Technology, 50:74-75.

55. Mohamed, P. G., Al-Saif, K. H. and Mohamed, H. (1999) Field evaluations of different multiphase flow measurement systems, Society of Petroleum Engineers Annual Technical Conference and Exhibition, Houston, TX, USA, SPE 56643 1:553-561.

56. Mokhtarzadeh-Dehghan, M. R. and Stephens, D. J. (1998) A numerical study of turbulent flow through a variable area orifice meter, International Journal of Computer Applications in Technology, 11:271-280.

57. Moore, P., Brown, G. J. and Stimpson, B. P. (2000) Ultrasonic transit-time flowmeters modelled with theoretical velocity profiles: methodology. Meas. Sci. Technol. 11:1801-1811.

58. Morrison, G. L. (1997) Flow field development downstream of two in plane elbows. ASME Fluids Engineering Division Summer Meeting FEDSM'97, Paper 3021.

59. Morrison, G. L., Terracina, D., Brewer, C. and Hall, K. R. (2001) Response of a slotted orifice meter to an air/water mixture, J. Flow Measurement and Instrumentation., 12(3): 175-80.

60. Morrow, T. B. (1997) Effects of flow conditioners on orifice meter installation errors, ASME Fluids Engineering Division Summer Meeting FEDSM'97, Paper 3006.

61. Niazi, A. and Gaskell, M. (2000) Building confidence with multi-path ultrasonic meters. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

62. Olivier, P. D. (1997) The effects of line pressure on the performance of thermal mass meters, International Instrumentation Symposium, Instrument Society of America, Aerospace Division, Orlando, Florida, 43:669-680.

63. Paik, J. S., Park, K. A. and Park, J. T. (1998) Inter-laboratory comparison of sonic nozzles at KRISS. FLOMEKO '98 Proceedings of the 9th International Conference on Flow Measurement, Lund, Sweden, 95-99.

64. Park, K. A., Choi, Y. M., Choi, H. M., Cha, T. S. and Yoon, B. H. (2001) The evaluation of critical pressure ratios of sonic nozzles at low Reynolds numbers. J. Flow Measurement and Instrumentation., 12(1):37-41.

65. Paton, R. (2000) Uncertainty analysis in the oil industry: an example based on the calibration of a pipe prover. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

66. Reader-Harris , M. J., Brunton, W. C. and Sattary, J. A. (1997) Installation effects on venturi tubes. ASME Fluids Engineering Division Summer Meeting FEDSM'97, Paper 3016.

67. Reader-Harris, M. J., Brunton, W. C., Gibson, J. J. Hodges, D. and Nicholson, I. G. (2001) Discharge coefficients of Venturi tubes with standard and non-standard convergent angles. J. Flow Measurement and Instrumentation., 12(2):135-45.

68. Riezebos, H. J., Mulder, J. P., Sloet, G. H. and Zwart, R. (2000) Whistling flow straighteners and their influence on US flow meter accuracy. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

69. Sanderson, M. L. (1994) Domestic water metering technology, J. Flow Meas. Instrum., 5: 107-13.

70. Skwarek, V. and Hans, V. (2000) The ultrasonic cross-correlation flowmeter ?new insights into the physical background. FLOMEKO'2000 the 10th International Conference on Flow Measurement, Salvador, Brazil: Paper B3.

71. Strzelecki, A., Gajan, P., Couput, J. P. and De Laharpe, V. (2000) Behaviour of venturi meters in two-phase flows. FLOMEKO'2000 the 10th International Conference on Flow Measurement, Salvador, Brazil: Paper D6.

72. Studzinski, W. and Karnik, U. (1997) Installation effects on orifice meter with no flow conditioner, ASME Fluids Engineering Division Summer Meeting FEDSM'97, Paper 3014.

73. Svedin, N., Stemme, E. and Stemme, G. (2001) A static turbine flow meter with a micromachined silicon torque sensor. Proceedings of the 14th IEEE International Conference on Micro Electro Mechanical Systems (MEMS 2001), Interlaken, Switzerland, 208-11.

74. Thorn, R., Johansen, G. A. and Hammer, E. A. (1997) Recent developments in three-phase flow measurement. Meas. Sci. Technol. 8:691-701.

75. Toral, H., Cai, S. and Akartuna, E. (2000) Multiphase flow rate identification by pattern recognition at Shell Auk Alpha Platform. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

76. Tuss, B. (1997) Wet gas multi-phase measurement. Offshore Technology Conference, Houston, TX, USA, 4:517-522.

77. van Bokhorst, E. and Peters, M. C. A. M. (2000) Impact of pulsation sources in pipe systems on multi-path ultrasonic flowmeters. North Sea Flow Measurement Workshop, National Engineering Laboratory, East Kilbride, Scotland.

78. Wee, A. (1999) Multiphase measurement system with fully redundant measurements to improve accuracy and simplify maintenance. 17th North Sea Flow Measurement Workshop, Paper 25 pp 346-361.

79. Worch, A. (1998) A clamp-on ultrasonic cross correlation flowmeter for one-phase flow, Meas. Sci. Technol., 9: 622-630.

80. Yeh, T. T. and Mattingly, G. E. (2000) Ultrasonic technology: prospects for improving flow measurements and standards. FLOMEKO'2000 the 10th International Conference on Flow Measurement, Salvador, Brazil: Paper A2.

81. Zanker, K. J. and Goodson, D. (2000) Qualification of a flow conditioning device according to the new API 14.3 procedure. . J. Flow Meas. Instrum., 11(2): 79-87

RECENT AND LIKELY DEVELOPMENTS IN FLOW MEASUREMENT

Roger C Baker, Institute for Manufacturing, University of Cambridge, UK

Summary

This paper attempts to summarise the considerable published work in flow measurement that has taken place over the past five or so years. The author has reviewed the literature⁵and brought together over 200 references. This task was to bring up to date the work in his Flow Measurement Handbook⁴. A large number of papers have been included, but many others have been missed, or omitted to keep the review within a reasonable length. In this paper an attempt is made to identify the most important trends in that period. The dominant factors are identified:

• management aspects such as application, cost of ownership, audit and maintenance;

• developments in meter design, and the dominance of three: the electromagnetic, ultrasonic and Coriolis;

• meters for wet gas and multiphase flows;

• industrial developments;

• market and manufacture.

In conclusion the paper suggests likely future trends:

• New types of flowmeter are likely to appear to complement the 100+ types already on offer, and manufacturers will attempt to offer a range for all applications.

• Multiphase flowmeters will continue to be developed for component flow rates, wet gas, and water-in-oil.

• Computational Fluid Dynamics (CFD) solutions might be capable of replacing rig testing in the near future.

• Previously insoluble applications problems are now within the scope of new materials.

• Modern developments in sensors and in signal processing may call for a review of existing meter designs.

• The power of digital methods to analyse signals is likely to affect meters increasingly, not least multiphase meters.

• Information technology will affect all stages from initial interpretation of the signal, through bus systems and interface to computers. It will also influence the modelling of the metered networks.

• Taking the whole manufacturing process from identification of product, through design to marketing, sales and product maintenance, we are likely to see much greater attention to reducing variation and waste, and increasing quality, over the next 10 years.

Because of limited space in this paper, the author has made a small selection from the vast literature, and the interested reader is referred to Baker^{4, 5}for more references. The author wishes to emphasise that the selection of papers, or of illustrations, should not be taken to reflect the value or otherwise of individual papers, or of particular devices and designs, but rather the author's need to provide examples of developments.

Management, metrology, flows, calibration

The management of flowmeters, at all stages from selection through application in complex systems, to identifying malfunction, is clearly an area where modern information technology methods would be attractive. How does one select? How do we allow for the costs of ownership? How can we check performance and identify emerging problems? It is an area where the manufacturers are moving forward in the development of CD-ROMs, and, presumably, will move on to e-communications.

Various approaches to error identification in systems have been reported. Menendez et al⁵¹used a model of a water supply net to deduce the errors in flow measurement from:

1. Analysis of the system;

2. Assignment to the meters of flow measurement uncertainty;

3. Estimate of flow distribution in the net;

4. Comparison of the estimated values with the measured values.

Others have used measures from independent instruments, and from billing data to identify malfunctions in, for instance, gas flowmeters.

There is a continuing concern with calibration and the requirements for traceabililty. Papers covered particular facilities, specific devices, very stable transfer standards and dynamic traceability. There continues to be a need for in situ calibrations. Le Brusquet and Oksman⁴⁴suggested combining computational fluid dynamics (CFD) predictions with tracer methods to obtain an improved prediction of the flow rate.

Paton⁶⁵discussed uncertainty analysis and made the point that the difficulty is in finding, recognising and judging the sources of uncertainty. He also touched on Monte Carlo methods which are beginning to cause interest.

Figure 1. NOVA flow conditioner (reproduced from Karnik³⁸with permission of the American Gas Association: Copyright, 1995 Operating Section Proceedings, Catalogue No. X59507, Figure 1, p 572).

Much work on flow conditioning has been reported recently. There may be problems due to flow-created noise for flow through conditioners when they are used with ultrasonic flowmeters. There are claims made that turbulent profiles can be recreated by plates, but possibly not the distribution and size of turbulent eddies. Karnik³⁸described the development of a Laws plate (Figure 1), which accounted for Reynolds number. For Re>10⁶, a single elbow and two elbows out-of-plane, 5D mixing length upstream of the flow conditioner and 8D downstream to the orifice plate, the meter error was within 0.1%. He also reported the sensitivity of the performance of perforated plate conditioners to the size of the holes.

Morrow⁶⁰compared results for tube bundle with cylindrical and hexagonal tubes, the Gallagher flow conditioner and no conditioner. The error for the Gallagher conditioner was less than 0.2 per cent for all β ratios from 0.2 to 0.75 for 4D to 12D between outlet of conditioner and orifice plate. Zanker and Goodson⁸¹noted that the new AGA 3/API 14.3 standard on concentric square edged orifice meters allowed other flow conditioners to be qualified for use by meeting certain type approval requirements. They mentioned tests such as: baseline, good flow conditions, two 90^oelbows in perpendicular planes, gate valve 50 per cent closed, 24^oswirl.

The development of CFD provides a tool for analysing flowmeter behaviour. Development continues^{36, 58}of solutions for real internal flows of the sort encountered in pipe flows by commercial flowmeters.

Flowmeter types and their development

Differential pressure (DP) flowmeters

Work on installation effects continues as shown by Studzinski and Karnik⁷². They commented that "the results of the tests indicate that the interaction of the orifice meter with various flow disturbances is a complex process and deviations of the orifice reading from baseline measurements are sometimes not as expected." The measurement of:

• C (Coefficient of Discharge),

• C with upstream fittings causing distortion to the inlet profile,

• C with upstream conditioners etc.

has been, and will continue to be, the subject of experimental programmes. The assessment of error which results from corrosion, wear (particularly on the inlet edge of the plate) and deposits on the plate, in the pipe and in the pressure tappings, is also important.

The effect of pulsation also continues to generate test data. McBrien⁵⁰reported tests at high gas pressures. For pressure in the range 46 to 51 bar and Re between 10⁶and 3 x 10⁶, the error was expected to be less than 0.5 per cent for rms to average pressure ratio ≤ 0.25. The error was expected to be less than 0.2 per cent for rms to average pressure ratio ≤ 0.1. Donoghue and Crane²⁴identified square root error and impulse line pulsation at CATS Terminal, Middlesbrough, UK.

The computation of the flow, particularly if coupled with upstream distortion, offers a considerable challenge and the results can be sensitive to the detailed application of computer programs. Mokhtarzadeh-Dehghan and Stephens⁵⁶presented computational analysis of fixed (Figure 2) and variable area orifice meters. For the fixed orifice they obtained discharge coefficients for corner taps about 1 per cent low and for D/D/2 taps about 0.8 per cent high compared with ISO5167, possibly due to modelling a zero thickness orifice. It would be interesting to explore this difference further, and to find the effect of changes in the turbulence model, edge treatment, upstream flow profile and turbulence distribution.

The renewed interest in the venturi meter is reflected in the number of papers. One particular area of interest is in high-pressure gas flows. Another, dealt with later, is in two- and multiphase flows. Reader-Harris et al⁶⁶gave installation distances for zero additional uncertainty for venturi meters, which appear to be at variance with the values in the ISO. More recently Reader-Harris et al⁶⁷reported tests for discharge

Figure 2. Computation of flow through an orifice meter: (a) velocity; (b) pressure; (c) turbulence energy; (d) wall pressure variation (reproduced from Mokhtarzadeh-Dehghan and Stephens⁵⁶with permission of the publisher: Interscience Enterprises Ltd, World Trade Center Building, Geneva 15, Switzerland CH-1215) .

coefficients of venturi tubes with non-standard convergent angles in gas and water and have discussed the change of discharge coefficient with time, due for instance to roughening of the tube. They gave an equation for the discharge coefficient that included a tapping term and used a Reynolds number related to the tapping diameter:

Re* = d_tapRe_d

The overall equation is then:

 1.0011 + 0.0123β - 0.0169e^{-0.4(Re*/105)}Re* > 60 000

C = 

 0.9878 + 0.0123β Re* ≤ 60 000

They compared this equation with the database for C obtained in gas and obtained an uncertainty of 1.23 per cent based on two standard deviations. For water the uncertainty found was 0.74 per cent. Their CFD work suggested that roughness in the venturi tube could account for 5 per cent or more error resulting in a decrease in discharge coefficient, while upstream roughness had smaller effects but tended to increase the coefficient. They also commented on the effects of deposits and wear on the coefficient.

The critical flow venturi nozzle appears to have the potential for very high precision. Paik et al⁶³gave data on discharge coefficients which suggest that the ISO uncertainty is too cautious. A special issue of The Journal of Flow Measurement and Instrumentation (Volume 11, Number 4, December 2000) was devoted to sonic nozzles. Papers dealt with numerical and experimental investigations, correlations for gases, uses as standards and provers, and various effects which influence the discharge coefficient. The editors of the issue, suggested¹⁰that the continued high level of interest in the critical (sonic) nozzle resulted from:

a) the increased use of natural gas, and the need to accurately monitor its flow;

b) advances in CFD allowing better modelling of the flows;

c) improved absolute gas flow measurement standards;

d) increasing importance of uncertainties in nozzle dimensions;

e) behaviour of very small nozzles.

Park et al⁶⁴evaluated the critical pressure ratios of sonic nozzles at low Reynolds numbers and developed a relationship between critical pressure ratio for small sonic nozzles and Reynolds number (based on throat diameter) given by:

P_cr= 0.9801 – 39.046 x Re_d^-0.5

They recommended this for Reynolds numbers below 10⁵, but recommended a safety factor also.

Morrison et al⁵⁹undertook further tests on a slotted orifice plate with an equivalent β ratio of 0.5 in an air and water flow. Their work suggests that it is insensitive to upstream flow conditioning and responds to two-phase flow in a well-behaved manner. Lawrence⁴³discussed the application to wellhead metering of the V-cone technology, and suggested that it should be considered for the solution of particular problems.

Positive displacement (PD) and turbine flowmeters

References continue in the literature on positive displacement meters of various sorts, but they tend, mainly, to confirm what we already know - that they are high precision, high reliability devices. Thus the diaphragm gas meter, the gear meter, helix-type meter and sliding vane meter and the rotary

piston positive displacement meter as a reference meter for gases, all get a mention, as well as a servo-controlled positive displacement flowmeter.

Mickan et al⁵³reported systematic investigation of flow profiles in pipes and their effects on turbine gas meter behaviour. Cheesewright et al¹⁴undertook field tests of correction procedures for turbine flowmeters in pulsatile flows. Large over-registration can occur in gas meters and, although the authors developed a software tool capable of estimating the true flow to about 1 per cent, there were many circumstances where this accuracy could not be maintained.

Svedin et al⁷³described a meter with a constrained turbine wheel (static turbine flowmeter) which converted volume flow into a torque. The device uses a micromachined silicon torque sensor. The optimum size was found to be a blade length of 2.7 mm with an angle of 30^o.

Other new designs with new sensing methods, large turndown ratios, self-checking or condition monitoring, are likely to be of particular value for metering high value fluids in fiscal and custody transfer applications.

Vortex flowmeters

The vortex flowmeter is susceptible to impulsive forces applied to the pipe, flow pulsation, inlet flow and probably turbulence level. Improving the design of the piezoelectric sensor will reduce the effects due to structural vibration, and new means of sensing may enhance range and performance.

The vortex meter's application to dry saturated steam may prove to be one of the most important contributions of this device. It would be interesting to explore means of sensing, by impact or by entrainment in the shed vortices, the amount of water in the flow and thus to obtain a dryness fraction for wet steam.

Electromagnetic flowmeter

Bates^{6, 7}seemed to suggest that with a smaller upstream pipe of 45 mm compared with 50 mm for the flowmeter bore, and with misalignment of 3 mm, the errors were of order 1 per cent. He has suggested that the error for a flowmeter downstream of a mitre-bend/reducer combination, when the electrode plane is about 7.7D downstream of the reducer outlet, can range upward from 1 per cent for higher flow rates to 2-4 per cent for lower flow rates.

Ultrasonic flowmeter

The dominant work on ultrasonic flowmeters has been:

• on ultimate performance particularly in respect of fiscal transfer;

• on the response of the meter with various paths to flow profile;

• on the practical effects of roughness on response;

• on extracting the maximum from the signals including neural network approaches.

The performance of these meters is continually improving. For fiscal applications with oil, repeatability appeared to be of the order of 0.01 per cent²⁰. Yeh and Mattingly⁸⁰have undertaken further tests on commercial meters and suggest that the zero flow and remove-replace performance has improved. They confirmed that the very high "dry calibration" expectations for multipath meters may be achievable with uncertainties of 0.2 per cent or better.

Lunde et al⁴⁸also noted that ultrasonic flow metering is now recognised as an alternative for fiscal metering of gas. However they also noted some challenges for improved accuracy and traceability including: transit time corrections, reduced sensitivity to operational factors and installation conditions, increased use and confidence in dry calibration. Niazi and Gaskell⁶¹commented that the uncertainty of the multipath ultrasonic meters they tested was better than 0.4 per cent when tested under ideal conditions, but confirmed the problems of flow-generated noise on ultrasonic meters.

Moore et al⁵⁷have taken theoretical profiles of distorted flows and applied them to ultrasonic meters with various path arrangements. They have demonstrated the possible sensitivity to flow profile of meters which, essentially, use diametral paths rather than half radius paths. The effect of flow profile appears to cause changes in measured flow compared with actual flow of up to 20 per cent for one path, and up to 3 per cent even for the four path version. It would be interesting to see the equivalent results for the half radius types.

Dane and Wilsack²¹investigated the influence of upstream pipe wall roughness on ultrasonic flow measurement. They found, over the range of conditions investigated, an increase in meter reading of 0.1 to 0.2 per cent for an increase in arithmetical mean roughness of about 5μm to about 20μm.

Lynnworth et al⁴⁹described the use of relatively high acoustic impedance transducers using solid piezoelectric materials for air and gas flows. They were capable of being subjected to temperature extremes, pressures from vacuum to 10 bar, vibration and shock, and the end faces can be contoured to create different distributions of sound wave. They may be capable of clamp-on crossflow measurement in plastic ducts. Liu et al⁴⁷described some buffer waveguides for use with hot liquids and appears to suggest that they can operate continuously with liquids up to 260^oC (Figure 3).

The problem of acoustic interference is causing errors in meters. Riezebos et al⁶⁸discussed whistling flow straighteners. It appears that high ultrasonic background and resulting vibration results in more frequent errors in measurement of path times. Pulsation sources in pipe systems also appear to have an impact on multi-path ultrasonic flowmeters⁷⁷. Sources range from fractions of a Hertz for process dynamics, through reciprocating machinery, flow induced pulsations, rotating machinery to valve noise of over 10 kHz. Substantial pulsation errors appear to be negative, for example with a pulsation amplitude of 8 per cent and a frequency of 25 Hz, the error in reading was between -3 per cent and -8 per cent.

Figure 3. Ultrasonic shear wave clamp-on buffer waveguides for hot liquids. (Copyright Panametrics, reproduced with permission).

Worch⁷⁹discussed a particular correlation flowmeter claiming a measurement error of less than 2 to 3 per cent for Reynolds number range from 250,000 down to 25,000 and even down to 4,000. Skwarek and Hans⁷⁰also reported work on cross-correlation meters and highlighted the different origin and treatment of amplitude and phase modulating events and the possible extension of application to gases.

Battye⁸discussed acoustic considerations affecting the design of demodulators for the ultrasonic correlation flowmeter. With transducers in contact with the water to minimise unwanted resonances, the standing wave patterns were, nevertheless, thought to contribute up to 2.5 per cent uncertainty in the flow reading without in situ calibration. With non-wetting transducers, the uncertainty might be larger. The author continued to see the meter as having a particular application to multiphase flows.

The ultrasonic signals offer more data than is usually exploited. One aspect is density to obtain mass flow measurement. Lansing⁴¹discussed smart monitoring and diagnosis for ultrasonic gas meters noting that the four common diagnostic features are speed of sound by path, path gain levels, percentage of accepted pulses and signal-to-noise ratio.

Thermal flowmeter

Olivier⁶²discussed the effects of line pressure on the performance of thermal mass meters. The results presented in the paper indicated a strong effect at low flow rates, while for higher flow rates the effect may have been only apparent above, say, 100 psig. The author noted that the effect varied between gases, and accounted for it on the basis of a modest effect of pressure on the heat transfer coefficient for forced convection, but a stronger one for free convection.

Ashauer et al¹mentioned four different strategies for thermal flow sensing: heat loss, heat addition, convection and time-of-flight. They presented some numerical work on the propagation of heat pulses in the flow, and described the design and construction of a micrometer which combined two strategies: the thermotransfer, and the time-of-flight. With these two it was possible to measure flows from 0.1 mm/s to 140 mm/s. An overlap between the two methods may allow calibration checking.

An important development is micro-thermal sensors that allow fabrication of air-flow sensors as well as signal-conditioning electronics on a single chip. With such standardised components the possibility of bypass metering with multiple and replaceable bypass meters would be an interesting development.

Coriolis flowmeter

As for ultrasonic flowmeters, this is an area of considerable development, and the developments centre in:

• single tube technology (Figure 4);

• improved modelling to take account of detailed geometry and conditions such as the effect of fluid pressure;

• experimental work to assess application effects;

• use with two-phase flows;

• new sensing and signal processing.

The Coriolis meter lends itself to the use of sophisticated computer models taking in the fine detail of the construction as well as allowing for fluid behaviour, compressibility, homogeneity and other features. Hulbert et al³⁷discussed a finite element numerical analysis of Coriolis flowmeters, and compared it with experiment. They noted that axial tension terms were important. Keita³⁹used a multipurpose finite element code with fluid structure interaction to obtain the characteristics of the Coriolis meter in the form of a straight pipe with clamped ends. His results compared well with previous experimental and theoretical values. The paper suggests that the procedures are far from straightforward as yet. Cunningham and Hensley¹⁹used IMAT and Matlab for Coriolis flowmeter design. Kutin and Bajsic⁴⁰discussed their work

(a)

(b)

Figure 4. Photographs of straight tube Coriolis flowmeters: a) Krohne (reproduced with permission of Krohne); b) Endress+Hauser (reproduced with permission of Endress + Hauser).

on stability boundary effect in Coriolis meters when the measurement approached the critical flow rate. Unfortunately the authors did not appear to give typical values.

Cheesewright and Clark¹³undertook theoretical and experimental work on the effect of flow pulsations on Coriolis mass flowmeters. Cheesewright et al¹⁵(cf. Reference 9) identified external factors which influence the calibration of Coriolis flowmeters. Pulsation and vibration near the Coriolis frequency, or the drive frequency, may cause errors. They concluded that almost all Coriolis meters were affected by vibrations at the Coriolis frequency. Other frequencies could also affect the calibration. Inlet flow conditions had little or no effect, but air/water flows introduced errors. The authors concluded that it might be possible to monitor the meter signals and generate warnings of most conditions apart from frequencies near the drive frequency.

Clarke¹⁶discussed non-linear control of the oscillation amplitude of a Coriolis mass flowmeter and proposed positive feedback of the output velocity to cancel the internal damping and described the method of determining the gain, and the insertion of an inverse non-linearity in the loop to create a linear system. The approach appeared to be valid and effective. Henry et al³⁴discussed a self-validating digital Coriolis mass flowmeter.

For gases there are various questions raised in the literature about the accuracy of a gas Coriolis meter. Effects due to compressibility and other features of the gas could mean that, for the highest accuracy, the meter becomes sensitive to type of gas.

For two or more components in the flow there are problems in the use of these meters due to the possibility that the components can move relative to each other.

A silicon resonant sensor structure for Coriolis mass-flow measurements of very small size has been described recently²⁷, and this may be an important development for the future.

Meters in and for multiphase flow

The predominant driving force in the development of multiphase flow measurement has been the oil industry, with its huge financial rewards, substantial capital investment, and its ever-changing target as the make-up of the flow from the oil well alters with time. The number of multiphase instruments in use may be of the order of 1000 or so, but comparatively few have been on the seabed. Perhaps the more immediate goal is for water-cut in difficult fractions, and wet-gas. Multiphase flow measurement is also required in other applications. The food industry is another obvious one interested in such flow measurement. Nevertheless, most papers derive from multiphase flowmeters applied to the oil industry.

One of the most interesting developments relates to the renewed interest in the Venturi as a means of measuring two-phase, including wet gas flows and multiphase flows³⁰. The oil and gas industry is showing particular interest. Strzelecki et al⁷¹tested a venturi meter with a beta ratio of 0.6 in an air-water facility. They found that their results fell between the Murdock and de Leeuw correlations, veering towards that of de Leeuw for increased air flow rates. Other areas under investigation are sensitivity to slip between phases and the effect of bubble size. Couput et al¹⁷(cf. Reference 18) reported on work in progress on the behaviour of a venturi meter in two-phase flows (wet gas). Their experimental results fell between the Murdock and De Leeuw correlations, and tend towards the De Leeuw prediction when air flow rate increases. Simulations reported showed droplet paths for three sizes of droplet, and showed, as expected, that the larger the droplet the less they followed the path of the gas.

Daniel et al²²described a venturi-based wet-gas meter with on-line gas mass fraction estimation and appeared to claim maximum uncertainty in gas mass flow rate over a 10:1 turndown of less than 3 per cent over a twelve year period.

Gopal and Jepson²⁹described the development of a novel non-intrusive, ultrasonic flowmeter for wet gas pipelines which has the capability of measuring liquid film thicknesses of order 0.1 mm and mist contents in gas flow of 4-5 per cent. Up to eight flush-mounted ultrasonic transducer pairs are used to obtain

transit time and attenuation of signals. It appears to be capable of handling stratified, annular and mist flows.

Thorn et al⁷⁴reviewed developments in three-phase flow measurement. They identified the oil industry's interest in three-phase meters as dating from the early 1980s when important developments were taking place, not least at Chr Michelsen Institute in Bergen where two of the authors were working. The authors suggested that a multiphase flowmeter (MPFM) would be capable of measuring each phase with an uncertainty of about 5 per cent, should be non-intrusive, reliable, flow regime independent, and suitable for the full component fraction range. The authors identified the main means of component fraction measurement as gamma-ray attenuation and electrical impedance methods. For component velocity measurement they mentioned cross-correlation and the use of the venturi meter for this purpose if the flow is well mixed. They provided a very useful overview of systems available or under development, and identified their technologies and the separation and homogenisation required for the systems. Hall et al³¹also reviewed the status of multiphase flow metering and provided an explanation of the various commercial meters and the means by which they measured velocity and phase fractions. They claimed that the minimum achievable uncertainty had reduced to around 10 per cent across much of the flow range.

The past five years or so have seen major advances in MPFM trials. In the mid-90s Hartley et al³³trailed a gamma-ray multiphase flowmeter developed by CSIRO on the West Kingfish oil platform. It used single and dual energy gammas and the tests were in a vertical line with upward flow. Average relative errors were:

• for liquids 3.9 per cent;

• for gas 7.6 per cent;

• for oil 7.9 per cent;

• for water 5.2 per cent.

Water cut was determined to 3.3 per cent. They note that for gas wells with high gas volume fractions, say above 95 per cent, the better approach to water cut may be using microwave methods.

There have been regular reports on instrument development and testing. Fluenta's MPFM used a capacitance sensor unit, an inductive sensor unit, a gamma densitometer, a venturi meter and pressure and temperature transmitters. The whole was mounted on a portable skid, with all the necessary electronics in an explosion-proof capsule. The inductive sensor was used where water was the continuous phase. Cross-correlation techniques between pairs of electrodes provided a measure of the velocities of gas and liquid. pVT information was obtained from an equation of state^{26, 45}. The earlier paper suggested that there were recommended orientation and upstream installation arrangements for the meter. Turndown was at least 10:1, with minimum multiphase velocities in the range 1.5 to 2.5 m/s depending on the amount of gas fraction. Velocities up to 15 m/s were acceptable. Pressures ranged from 108 to 128 psia (7 to 8 bar) and temperature from 126^oF to 170^oF (52^oC to 76^oC). Compared to the reference test separator the MPFM agreed to within about 10 per cent on fluid and gas production, and water cuts were within about 10 per cent absolute of wellhead water-cuts. These results were for gas volume fractions of 93 to 98 per cent. The meter could obtain the instantaneous composition of the flow. Caetano et al¹¹discussed the operation of the MPFM at the Albacora field off Brazil. The authors appeared to claim about 5 per cent or better agreement between the meter and the separator for accumulated volume.

The Framo meter, as described⁴⁶, consisted of three primary components:

i) a static flow mixer to homogenise the flow and eliminate flow regime effects;

ii) a venturi section to measure the composite or bulk flow through the meter;

iii) a gamma ray absorption meter embedded in the throat of the venturi.

The dual energy gamma meter provided fractions of oil, water and gas. There was also a means of determining salinity from the gamma spectrum. The authors noted flow regimes and salinity as two of the most important environmental effects on these meters. Tuss⁷⁶described the Framo/Daniel meter, which used an ejector tube to mix the high gas volume flows, and has been tested in the North Sea. He also discussed other meters such as the Agar which used a combination of positive displacement, venturi and water-cut meters. Atkinson et al²described cooperation by Schlumberger and Framo Engineering AS to produce a new MPFM, the VenturiX, consisting of a venturi and a dual energy composition meter at the throat. Most recently Larson⁴²described operational experience with a subsea multiphase Framo flowmeter in the West Brae Field. The meter was based on an insert design, allowing for easy retrieval of all the critical items: electronics, transducer, gamma isotope, detector, venturi and ROV-mateable connector for power, signal and flushing media.

Wee⁷⁸discussed a multiphase measurement system with fully redundant measurements to improve accuracy and simplify maintenance. The system components were: MFI multiphase composition meter, MFI multiphase cross-correlation velocity meter, venturi, dual temperature, pressure and differential pressure transmitters, PVT module and a multiphase management system. The first two are in a compact straight spool piece. The meter is, therefore, claimed to have redundancy in measurements. The liquid flow rate is claimed to be in the range 6 to 10 per cent at 90 per cent confidence level, the gas velocity as 10 per cent, water cut of 5 per cent or better and GVF of 2 per cent or better. Cellos¹²reported on a multiphase flow measurement system in Prudhoe Bay for high GOR applications by including partial separation of components. The GOR went up to 80,000 scf/bbl, with liquid flow rates of 100 to 15,000 bbl/day. Oil flow rate was obtained within 5 per cent and gas flow rate within 2 to 3 per cent. The system consisted, on the predominantly liquid leg, of an MFI multiphase meter having a 50 mm spool piece with no pressure drop and, on the predominantly gas leg, of a Coriolis meter. Microwave was used to measure the dielectric properties with the frequency inversely proportional to the square root of the mixture dielectric constant. Cs137 was used for the gamma densitometer to obtain the density of the mixture. The meter appeared able to cope with 0 to 100 per cent water cut. A microwave cross-correlation arrangement was used for flow rate. In addition a venturi could be added.

Others have proposed the use of non-radioactive methods. Nuclear magnetic resonance (NMR) has been proposed, but velocity resolved spectroscopy (VRS) was seen as a more distant possibility. Other systems have been described and some tested, and other applications considered such as viscous flows and heavy oils. Mohamed and Al-Saif⁵⁴described an MPFM that used a fluidic-flow-diverter to separate gas and liquid at high gas flow rates. The gas flow was measured using a volumetric meter while a positive displacement meter measured the gas/liquid stream. A venturi system measured pressure at four locations to solve the momentum equations without fluid property input, and a proprietary microwave oil/water meter was used for the water cut. Average error in liquid- and gas-production rates over 32 tests was about 6 per cent and 7.5 per cent respectively. Mohamed et al⁵⁵tested two MPFMs. One used a separator for the gas, which was then measured by vortex flowmeter, while liquid was measured with Coriolis and water cut by sampling, electrical and differential pressure methods. The other used positive displacement and venturi meters and microwave, and had a vortex bypass. They discussed foaming and increasing back pressure from the MPFM on well production.

Harrison et al³²discussed the effects of salinity variation on dual energy multiphase flow measurements and Mixmeter (Figure 5) homogeniser performance in high gas and high viscosity operation. They also noted that a third gamma energy could reduce errors due to salinity changes. de Carvalho and Antunes²³reported on new data from the BP/ISA Controls Multi-stream meter. Dyakowski²⁵discussed the application of process tomography and Toral et al⁷⁵described continuing work on multiphase flow rate identification by pattern recognition. Mi et al⁵²described a flow regime identification methodology with neural networks and two-phase flow models.

Figure 5. Photograph of a multiphase meter (reproduced with permission of Jiskoot Autocontrols)

Conclusions on the past 10 years and thoughts for the future

Ginesi²⁸in 1997 quoted Greg McCall of McCrometer as saying that environmental compliance and regulations continued to drive flow measurement in areas of the plant that traditionally have been unmonitored, or at best, poorly monitored. Applications like steam measurement on the utility side of the plant were also driving flow measurement. Other points Ginesi highlighted were:

• the extension of the application range of existing flowmeter types, e.g. vortex to lower Re;

• non-invasive designs;

• reliable measurements of multiphase flows – slurries, air in liquid, solid in liquid, gases with entrained liquids, mists and wet steam;

• more parameters measurable with one instrument at the same price;

• the need for an expert system for selection, to simplify the problem.

He noted the increasing accuracy of DP transmitters, the introduction by some companies of metal tube VA meters, improved life expectancy for turbines, versatile magnetic flowmeters for partially filled pipes and new designs of electromagnetic flow sensors inserted into the unlined steel tube.

The last 10 years have seen major developments in the measurement of water and gas for domestic purposes. The fluidic flowmeter described by Sanderson⁶⁹is an elegant solution to the need for a wide-ranging no-moving-part domestic water meter, and the ultrasonic meter for gases is another significant development.

The electromagnetic flowmeter has become a very reliable instrument, and with it the ultrasonic and Coriolis meters have moved from being useful to dominating the scene.

10 years ago this author³considered that future developments seemed likely to lie in the areas of:

a) non-intrusive, non-invasive, clamp-on, particularly ultrasonic techniques;

b) optical sensing methods with existing devices;

c) intelligent, smart, self-monitoring and monitoring the plant performance;

d) multiphase flow measurement.

Apart from (b) this was a reasonable prediction. There appear now to be, perhaps, eight main trends for the future:

Flowmeter types

While new types of flowmeter are likely to appear to complement the 100+ types already on offer, the interesting factors are that:

• individual manufacturers attempt to offer a range for all applications;

• electromagnetic, ultrasonic and Coriolis dominate the scene;

• interesting and useful CFD analysis is now possible, and may be particularly relevant to differential pressure devices, two-phase flow modelling in various meters and installation effects.

However, development can be expected in other flowmeters. Examples are the exploitation of information in the ultrasonic signals, and the optimisation of various designs of thermal flowmeter for improved performance.

There may be the possibility of an electromagnetic industrial design for hydrocarbon flows, and recently the first ultrasonic clamp-on flowmeter for gas has been launched.

The systematic analysis of the Coriolis meter, including computational modelling of the tube in its vibration, stressing and manufacturing variation, is an exciting development leading to a new generation of instrument. What will be the next major development? The single straight tube has been introduced by several manufacturers, and its evolution into an instrument insensitive to installation and environment will be highly significant. Could the next stage be shorter pipes, or multiple sensors? Will there be new vibrating elements within tubes or travelling wave devices? Can the manufacturers reduce the price of these instruments substantially?

In addition a number of micro devices are being developed, such as:

• the resonating micro-bridge mass flow sensor;

• a very small orifice with a nominal diameter of 100 μm and operating in the molecular flow regime for applications in ultra-high-vacuum technology;

• a silicon sensor depending on convective heat transfer resulting from the liquid flow, and made using an industrial bipolar process and micro-machining;

• a multisensor which includes an array of pressure, temperature and shear stress sensors.

Multiphase flow measurement

The development of the oil and gas fields and the changing nature and ratios of the components from the wells, has resulted in a need for developing solutions to keep up with these changes. The main instruments required appear to be for: component flow rates, wet gas, and water-in-oil. The financial importance of these devices, and the funding available for their development, has resulted in a large number of major programmes to meet the needs.

It would be satisfactory to have a flowmeter that could interpret all the variables occurring in a multi-component flow. Such a device might identify the position on a flow pattern chart for well-documented fluids and the mass flow of each component. To achieve this for gases and liquids, without errors due to water, sand, and wax, is difficult and most methods will need to compromise.

The wider uses of multiphase flowmeters, should also be remembered particularly in the food industry.

Computational fluid dynamics (CFD)

The advances in computer power, and the development of powerful computer codes, may soon make computational methods an alternative to testing. Hilgenstock and Ernst³⁵compared CFD (Fluent code) predictions with experimental measurements and found good agreement. They considered that the CFD solutions could provide more detailed information, could be as accurate as rig testing, and might be capable of replacing testing in the near future. Areas of particular concern relate to:

• detailed investigation of the flow in orifice and venturi, and the modelling of sharp edges and pressure tappings;

• prediction of the effects of installation, and particularly the use of perturbation methods;

• design and prediction of other meters such as: critical venturi, averaging pitot, turbine, and thermal flowmeter.

Materials

Advances in materials technology have meant that previously insoluble applications problems, often resulting in unsatisfactory compromise, are now within the scope of new materials. This applies to new developments in meters such as Coriolis, as well as meter covers etc. for many long-standing designs and instruments.

Sensor technologies

Sensors depend on vibrating elements, capacitance, resistance, voltage, magnetism, RF, optics, temperature, strain etc. Modern developments in these and in signal processing call for a review of these methods.

Digital processing

The power of digital methods to analyse signals and identify modes within them, is being applied to the Coriolis meter, but will also be important for other types. The idea that the meter signal contains data which can reveal information about the meter and the process is an important insight. However, if this is to be fully realised, it is essential that models linking the flowmeter behaviour with the noise characteristics be developed, to give a firm theoretical basis.

The technique for obtaining flow and wetness information from DP devices with different characteristics is likely to lend itself to the use of digital analysis of the signals. Signal analysis and neural network techniques are, also, likely to play a major role in future development of multiphase meters.

Information technology

This has already been assumed in several of the earlier development categories. It ranges from the initial interpretation of the signal and its conversion into digital or other form, the transmission and further processing, the linking with a bus system, the interface between the control computers and the people operating the system, and the modelling of the operation of flowing networks.

These and possibly other areas offer possibilities for new partnerships between the science base and the industry, which should be fruitful in developing a new generation of instrumentation.

Manufacture

Taking the whole process from identification of product need, through design and manufacture, to marketing, sales and product maintenance, the next 10 years are likely to see much greater attention to all aspects in an attempt to reduce waste and product failure, to increase profits and to attract larger markets.

References