水管理

掌握深水墨西哥湾生产水管理:25 年洞察 — 第 2 部分

本文是墨西哥湾采出水管理系列文章的第二篇,涵盖四个主题:设备、工艺配置、操作和流出物质量。

大片水域中的海上石油钻井平台
对于许多设计决策,该行业缺乏系统的、系统的方法来处理和管理生产水。
HeliRy/Getty Images

本文是系列文章的第二部分,涵盖四个主题:设备、工艺配置、操作和流出物质量。

第 1 部分包括介绍、生产水处理系统设计的六大主题的简要描述,以及对两个主题的更详细讨论:进料生产水的特性和化学处理。

本文包括第 1 部分的摘要,本文的结论部分涉及第 1 部分和第 2 部分。

抽象的

本系列由两部分组成,专门针对墨西哥湾 (GOM) 深水平台的采出水系统提供系统设计和操作指南。重点介绍了六个要素或主题:流体特性、化学处理、设备、工艺配置、操作和流出物质量。流入采出水的特性以及目标流出物质量决定了水处理的挑战。

由于深水环境中空间和重量成本高昂,水处理必须具有高强度,并且必须与周围设备高度集成。高强度是指设备能够管理高体积流量、短停留时间,并且占用空间小。高度集成的系统设计概念是指将水处理设备、工艺配置和化学处理集成到一个高性能单一系统中的系统工艺设计,该系统可去除极高比例的污染物(主要是悬浮在水中的油)。

充分利用一切机会,确保采出水得到高标准的处理。调峰、界面排污和作为澄清器的断流罐只是改善采出水质量的几个机会。

设备

提供基本信息,帮助设计团队选择处理采出水的设备,以满足舷外水处理的监管准则。这包括初级和次级除油设备——水力旋流器和浮选机(Walsh、Arnold 和 Stewart)。

初级分离设备(分离器、水力旋流器)取决于密度差、污染物(液滴和颗粒)大小和流体粘度。二级(浮选)性能取决于污染物大小、界面吸附(取决于污染物化学和化学处理)、气泡大小和气泡的总表面积(气水比)。三级分离设备(如过滤器、膜等)取决于污染物(絮凝物)大小、介质表面特性(取决于污染物化学和化学处理)、介质表面积、表观速度以及反冲洗等细节。

在墨西哥湾深水区,设计目标主要通过定制设计实现,主要是水力旋流器和浮选。这些技术被认为是“常规”技术,因为它们已经存在了几十年。然而,这些常规技术的详细设计特点并不为人所知,只有通过事后评估和对整个地区最佳实践的考察才得以揭示。

对于水力旋流器,需要入口处有更高的压力(约150 psig),以便更灵活地调整压差比和排出比等操作参数,从而实现最佳衬管效率。

定制设计特点包括小直径衬管(用于更大的前向压降和分离效率)、已成功使用的吞吐量分级(0.25/0.5/0.75)和大型废料口(用于削峰)。这些特点可以融入众所周知的传统水力旋流器中。

对于浮选,已采用多种方法来满足设计目标。浮选设备的定制包括产生大量小直径气泡的方法;设备设计的级数;浮选化学品的注入/在线混合;以及阀门和管道流动设计以最大限度地减少剪切。通过性能分析,发现单级浮选设计不适合深水应用。

本文将不讨论三级设备的分离效率,因为它通常不需要满足船外监管水质规范,并且通常不用于海上作业。

1.水力旋流器——初级除油设备

水力旋流器被视为海上作业中从采出水中去除分散油滴的“标准技术”,并且已在陆上成功使用。水力旋流器的除油效率是入口流和出口流中的油水 (OiW) 水平之差除以入口流中的 OiW(Walsh、Arnold 和 Stewart)。

1.1 分离的驱动力

水力旋流器利用衬管产生高离心力(1,000-2,000 G,其中 G 是重力常数)以根据密度对流入的油/水流进行分类。

通过使衬管内的流体产生旋涡运动,可以产生很大的离心力。旋涡运动产生的角加速度增强了分散的油滴和水之间的密度差的影响。油比水轻,因此会迁移到衬管的轴向中心。由于进料和弃料口之间的压力差,油核的流动方向与水的流出方向相反。如图 1.1所示,油被导向弃料口(溢流),水通过流出口(图 1.1 中的底流)流出。弃料核从旋风衬管顶部的弃料口中抽出。

应用改进的斯托克斯定律方程来分离油滴和水。

V 0 = Fc * 1.78 * 10 -6 (Dr) d 2

其中Fc是离心力,重力的倍数,V 0是油滴的垂直速度(单位为英尺/秒),d是油滴直径(单位为微米),A是水和油之间的比重差,μ是流体粘度(单位为厘泊)。油滴向中心核(或轴)的迁移速率由斯托克斯定律中的相同变量控制,但有一个例外。斯托克斯定律以重力为驱动力,而在水力旋流器中,旋流运动的离心加速度提供驱动力。

各种衬管部分

衬管有四个部分:圆柱形旋流室、同心缩径段、细锥形段和圆柱形尾段(图 1.1)。同心缩径段和锥形段之间的位置是旋流器尺寸规格(水力旋流器直径)的点。流体加速(切向速度)发生在缩径段、锥形段和平行段中。缩径段和锥形段为圆锥形状,平行段为圆柱形状。

带有标记部分的单个气旋示意图
图 1.1 — 单个旋风分离器的示意图,标有部分,显示典型除油旋风分离器的几何形状和流体路径。这种旋风分离器是封装除油水力旋风分离器单元中使用的典型旋风分离器。

内衬包装

大多数海上水力旋流器装置都有一个容纳许多衬管的压力容器(图 1.2)。最常见的成套装置有一个进料流和一个产品流和废弃流。
 

将多个旋风衬套封装在一个压力容器内。 
图 1.2 — 在单个压力容器内封装多个旋风衬套。该剖面图显示了废料集管和废料集管下方的进料部分。图中详细显示了单个衬套,包括将衬套密封在废料集管中的 O 形环。

包装是水力旋流器安装的一个重要方面。制造商开发了多种包装,具有各种理想的特性。设计团队需要审查各种包装,以选择最适合设计项目的包装。

在每个压力容器中合理选择衬管数量(例如 50 到 60 个)非常重要。根据操作和维护方面的考虑,选择每个压力壳体中的衬管数量以处理大约 25% 的总产水率或最多约 50,000 BWPD 可能是有用的。

当其中一个带衬里的容器因维修而停止使用时,提供一个备用压力容器可能会很有优势,尤其是在海上作业中。

1.2 工作原理

1.2.1 分离、产能和设备成本

分离效率在很大程度上取决于衬管的直径。直径越小,衬管的吞吐能力越低,但旋流速度越高,油水分离的驱动力越大。容量越低意味着给定水流量需要更多的衬管,因此设备成本会更高。

在深水作业中,由于剪切力过大,且长管路中液滴聚结的空间有限,液滴尺寸非常小。基于此,直径较小(高效)的衬管是首选。这是选择衬管时需要做出的基本决定之一。

图 1.3中的曲线是根据特定密度差和水粘度的经验相关性针对不同衬管尺寸生成的。这些变量的差异会改变结果,但不会改变整体结论。

图1.3——各种水力旋流器设计的典型分离效率曲线。 
图1.3——各种水力旋流器设计的典型分离效率曲线。 

对于深水海上作业,最好使用直径较小的水力旋流器(例如10毫米或25毫米)来去除小直径(<20微米)分散的油滴。

1.2.2 流量和水力旋流器分离效率(工作范围)

调节比。如图 1.4所示,水力旋流器的分离效率通常随着流速的增加而增加,并在很宽的流速范围内保持不变。在一定的高流速下,分离效率开始迅速下降。每个制造商的水力旋流器设计都有其独特的曲线。图 1.4 中的曲线显示调节比约为 4 比 1。出于操作目的,最好以最大额定流速的 80% 使用水力旋流器,以处理操作中通常观察到的液塞引起的流体速率波动。

图 1.4——在很宽的流速范围内分离效率与流速的关系。
图 1.4——在很宽的流速范围内分离效率与流速的关系。

压力

三种压力对于水力旋流器中流体流动和分离效率的优化非常重要:(a)前向或入口压力,(b)处理过的流出物或底流压力,以及(c)排出物或溢流压力。

底流压力是入口压力和通过水力旋流器的液体体积流速的函数,这会产生压力下降。

在典型的安装中,水力旋流器与分离容器相连,分离容器的液位控制阀位于水力旋流器的下游。因此,入口压力是容器内的压力,底流压力成为上游压力,用于确定容器液位控制阀的尺寸。

入口压力和底流压力之间的差值称为前向压降。前向压降的重要性怎么强调也不为过,但它经常被错误地指定。应最大化该压降,因为它提供了分离的驱动力。入口压力由上游分离器或适当(低剪切)选择的泵提供,是三个压力中最高的。出于设计目的,最好使该进料压力大于约 125 psig 或最好约为 150 psig。高入口压力将流体推过衬管,产生高离心力。

流体芯中的低压允许油水反向流向弃流口(弃流孔)。弃流口是一个非常小的孔,直径约为 1-3 毫米。这种设计是为了确保水力旋流器的弃流率很小,并且只有来自中心油芯的流体沿溢流方向流动。

溢流压力由外部阀门和阀门控制系统控制。

拒收率。拒收率或溢流率与入口流量之比称为拒收率。它通过调节溢流和底流排放管道上的阀门来调节。

拒绝率 = (溢流率/入口流速) * 100

水力旋流器系统具有最佳拒收率。系统在低于最佳拒收率的情况下运行将导致除油效率降低。安全裕度可以通过量化拒收率函数中的去除效率下降来确定。在大多数情况下,拒收率约为 2.5。

含油废弃物流中的分散颗粒可能导致小废弃物孔(约 2 毫米)堵塞并影响除油效率。这可以通过定期反冲洗废弃物孔来纠正。

压差比 (PDR)

PDR = (P进水— P排出水) / (P进水— P出水水量)

P进水口为进水口压力,P出水口为出水口压力或 P溢流口,P出水口   为出水口压力或 P底流口。对于油水分离,最佳 PDR 在 1.7 至 2.0 之间,确切的 PDR 需要在运行过程中确定。

PDR 必须大于 1,才能迫使核心中的油反向流动。对于给定的水力旋流器几何形状,PDR 控制被排出的流体量。图 1.5深入了解了 PDR、排出流量和分离效率之间的关系。右图是排出率和 PDR 之间的关系。图 1.5 显示了两种特定水力旋流器设计之间的关系。

图 1.5——两张图共同展示了 DP 比 (PDR) 和分离效率之间的影响。这两个变量通过截留率联系在一起。
图 1.5——两张图共同展示了 DP 比 (PDR) 和分离效率之间的影响。这两个变量通过截留率联系在一起。

性能数据。图 1.6显示了 2004 年 Auger 平台上水力旋流器性能的现场数据。数据显示,水力旋流器的主要优点是显著降低下游的油浓度(调峰)。这可以显著提高下游水处理设备(如浮选槽)的性能。水力旋流器性能(方块)从 2004 年 10 月开始逐渐下降,这是由于反冲洗频率降低和流速降低,而没有调整衬管数量。

现场数据。FWKO 排放口处用红外测量的油浓度
图 1.6 — 现场数据。用红外测量 FWKO 排放口(菱形,水力旋流器进料)和水力旋流器排放/浮选进料(正方形)处的油浓度。请注意,FWKO 排放口(水力旋流器进料)达到非常高的值。水力旋流器的主要优点是“削峰”。这可以提高浮选性能。

1.3. 水力旋流器的优点和缺点

1.3.1 优点

  • 调峰
  • 流体停留时间约2秒
  • 灵活的模块化设计。
  • 易于操作和维护。无活动部件。
  • 在大多数情况下,能够非常有效地去除大于 15 微米的油滴
  • 广泛且可预测的操作范围
  • 对平台运动不敏感
  • 对撞击不敏感
  • 快速启动和恢复
  • 油水拒收率约2%

1.3.2 缺点

  • 水力旋流器系统效率对液滴尺寸分布的敏感性
  • 需要相对较高的压力(约 150 psig)才能实现最佳水力旋流器性能
  • 需要低剪切泵来提高低压系统的工作压力
  • 小直径的废液口容易堵塞,因此需要定期反冲洗以尽量减少堵塞的可能性
  • 存在油湿固体时性能会下降

2.浮选机—二次除油设备

在本节中,我们简要介绍了设计工程师在为特定应用选择浮选设备时应考虑的基本信息。我们不描述各种浮选设备的细节,而只是提供有助于设计工程师选择设备的指导原则。

浮选是次要工序。如果系统中有水力旋流器,浮选总是位于水力旋流器的下游。

浮选通常能够将少量油性固体与分散的油颗粒分离,从而提高系统的效率。与液-液旋流器相比,气浮设备在分散油去除方面通常更有效,因为它能够去除分散的油滴和至少一些油性固体。

浮选装置利用细小分散的气泡将分散的油滴和油性固体带到表面。油滴和油性固体附着在气泡表面,并以根据斯托克斯定律估算的速度上升到表面。分散气泡附着在油滴和油性固体上会降低比重,从而产生更快的上升速度并加速其分离。

使用化学药剂将分散的液滴凝结、絮凝,气泡与絮凝后的液滴在水面形成泡沫层,通过撇取的方式将水面上的污染物除去。

2.1 分离机理

在浮选槽中,气泡被注入或诱导。气泡迅速上升并与油滴碰撞。碰撞频率取决于油滴浓度、气泡浓度以及油滴和气泡的投影面积。极少数碰撞会导致气泡捕获油滴。这被称为“捕获效率”。这大约是 1/1000。值得注意的是,与使用小气泡时的捕获效率相比,使用大气泡时的捕获效率要小得多。捕获效率取决于油和气的表面化学性质。它受凝结剂和絮凝剂等化学添加剂的强烈影响。

图2.1——气泡上升(蓝点)和油滴的示意图。显示了两个气泡的碰撞路径。
图2.1——气泡上升(蓝点)和油滴的示意图。显示了两个气泡的碰撞路径。
图2.2——碰撞频率与捕获效率。
图2.2——碰撞频率与捕获效率。

图 2.2是显示气水比和气泡大小对碰撞频率和捕获效率影响的示意图。为了更好地从浮选槽中的采出水中分离分散的油滴,需要小气泡和高气水油比。

2.2 设计考虑

浮选是一种接触分离过程。它依赖于气泡与油滴和固体颗粒的接触。重要的变量是:

  • 气水比(体积比)
  • 气泡大小(直径)
  • 分级(单级、双级、多级)
  • 化学应用(表面化学),如凝结剂和絮凝剂
  • 污染物大小和表面化学

设计变量包括较大的污染物油滴和油性固体颗粒、整个浮选槽中均匀的气泡分布、高的气水比和小(最佳尺寸)的气泡。应尽量减小气泡尺寸,同时确保气泡足够大以迁移到油/水界面。可以通过使用凝结和絮凝等化学药剂来增加污染物尺寸。这些化学品注入水力旋流器的下游但在浮选槽的上游。在浮选槽中使用多个阶段是提高分离效率的一种非常有效的方法。例如,在大多数应用中,水平液压和机械诱导浮选槽通常使用四个分离阶段。浮选槽中的总流体停留时间约为 5 分钟。
除了上述变量之外,机械设计中还有一些重要的实际考虑因素。

  • 必须以不干扰浮选过程的方式引入油水。
  • 湿油和清洁废水的排放方式必须不会对进料造成污染,也不会干扰工艺过程。
  • 整体重量和空间必须尽可能低,同时实现较高的分离性能。
  • 浮选槽中的废液必须得到妥善处理(例如,在设计合理的污水池中),以确保对整个采出水处理系统的影响最小。多级浮选装置中的废液通常占浮选槽流入流量的 5% 至 10%。

与大多数水处理设备一样,将浮选装置集成到设施或工艺中对于成功运行至关重要。集成最重要的方面是处理废渣。

2.3 浮选设备

上游石油和天然气工业中使用的气体浮选槽有四种类型(Walsh Arnold 和 Stewart)。

  • 溶解气浮选
  • 水平多级水力诱导浮选
  • 水平多级机械诱导浮选
  • 垂直诱导气浮——常用于深水作业

几乎所有卧式装置都是多级的。我们不提供任何有关设备的详细信息,但只会强调一些功能,这些功能可以帮助设计工程师为应用选择最合适的装置。

2.3.1 溶解气浮选(DGF)装置

在 DGF 中,气体在高压(约 20-40 psig)下溶解到循环的清洁水流中(约 20-25%)。天然气通常用于排除氧气。饱和气体的水随后通过阀门进入浮选室。当水流进入浮选室时,溶解气体会从溶液中以小直径气泡(20 至 60 微米)的形式释放出来。这些气泡比典型的诱导气体装置中的气泡小得多,后者大约大 5 到 10 倍。溶解气体通常为 0.2 到 0.5 scf/bbl 待处理水。

由于两个因素,可以注入 DGF 的气体量受到限制。第一个因素是气体溶解度的热力学限制,这取决于水的温度和气体的压力。第二个因素是可以回收多少处理过的水来将气体注入给水的实际限制。虽然 DGF 中会产生小气泡,但由于注入气体的体积限制,气泡总数相对较少。因此,在大多数情况下,总分离效率约为 50%。DGF 可以是单级或多级单元。

由于分离效率低,DGF 在石油和天然气上游行业并不常用。

2.3.2 卧式水力诱导浮选装置

在这些装置中,气体分散是通过在每个槽段中循环一部分(约 25%)清洁废水来实现的。这些水通过文丘里管重新进入浮选槽,文丘里管从槽中水上方的蒸汽空间中抽取气体,并将其与水一起排放到槽底部。这些液压装置大多为四个槽。

该装置的容量必须设计为处理进料(未处理的流入水)和再循环水。

2.3.3 水平机械诱导浮选装置(例如 Wemco)

这些装置通过电动机转动的转子实现气体分散。转子转动时,它充当泵,迫使水通过分散器并产生真空,将气体从气层吸入立管并将其与水混合。当气水混合物通过分散器时,会形成小气泡。这样形成的气泡会聚集油性絮凝物并将其带到表面,然后撇去。一般而言,机械诱导气浮槽的效率高于液压诱导气浮槽。但是,由于多个电机和转子导致维护成本较高,因此液压装置通常比机械装置更受青睐。

2.3.4 分级和去除效率

如前所述,水平浮选装置通常是多级的。在四室浮选装置中,总流体停留时间通常为 5 分钟。大量水流通过底流挡板从一个室连续流向另一个室。从一个室流向另一个室的水流基本上是活塞流模式。

长期以来,浮选动力学分析一直用于预测 IGF(诱导气浮选)工艺中的污染物去除效率。根据该分析,效率是细胞停留时间和细胞数量的直接函数(Movafagian 等人)。图 2.3基于动力学分析。它表明,在单个细胞单元中,即使流体停留时间增加到 2.5 分钟,最大污染物去除效率也约为 70%。设计工程师应考虑到这一点。

图2.3——污染物去除效率与细胞停留时间和细胞数量的关系。
图2.3——污染物去除效率与细胞停留时间和细胞数量的关系。

2.3.4 垂直诱导气浮选装置(VIGF)

VIGF(通常为单级)最初设计用于解决波浪运动对深水浮动结构的影响问题,以减少占地面积,并作为上述大型多级水平装置的低成本替代品。VIGF 表现良好;然而,单船应用容易发生故障,不提供冗余或防止机械故障的保护。

当需要垂直浮选装置来满足系统设计或空间限制时,应考虑串联使用两个或多个容器,因为单槽装置的效率较低,如图 2.3 所示。目前,垂直 IGF 的主要优势是成本低、空间要求低、重量轻。

典型 VIGF 中的气泡是通过使用喷射器(如 Unicell 中的液压诱导)产生的,或者在某些设计中,使用专用泵来产生气泡。然而,在所有这些 VIGF 中,都使用单级,流体停留时间在 2 到 5 分钟之间。

垂直配置适合使用容器内部元件,以促进气体均匀分布、气泡/油滴良好接触以及油滴聚结。这导致了复杂的内部元件,这使得这些装置与传统的水平 IGF 装置有显著不同。使用聚结元件的一个主要缺点是它们容易被采出水中的悬浮固体堵塞。

2.4 浮选机的一般性能指南

  • 四单元浮选机的效率高于单个单元浮选机的效率。
  • 诱导气浮选装置的效率比溶解气浮选装置的效率高。
  • 经验证明,机械诱导浮选装置的效率比液压诱导浮选装置的效率更高。
  • 一般来说,大于 10 微米的油滴可以通过浮选装置有效去除。
  • 诱导气体浮选装置的正确设计和操作应能够去除约 90% 的分散油。

2.5 浮选机的优点和缺点

优点

  • 具有悠久使用历史的著名工艺
  • 功能并不完全取决于油相和水相之间的密度差异
  • 良好的调节性能(停留时间更长,分离效果更好)
  • 去除分散油的有效系统,尤其是与化学添加剂(凝结剂和絮凝剂)一起使用时
  • 一般情况下,如果操作和维护得当,并与设计合理的主要除油设备配合使用,可使舷外水中的分散油浓度低于 29 mg/L
  • 有效去除大于 10 微米的油滴

缺点

  • 流体停留时间较长(约5分钟),设备较重、笨重。
  • 不适合浮式生产系统。
  • 需要稳定持续的供给。易受油的冲击负荷影响,随后恢复缓慢。
  • 容易受到由于单个油井启动/关闭而引起的流量和水成分变化的影响。
  • 浮选槽中残留的油性絮凝物很难清除,这会降低出水水质,并导致受纳水体表面出现光泽。
  • 一般来说,对化学物质及其剂量非常敏感。
  • 操作和维护需要大量人力。
  • 油滴的剪切可能发生在液压诱导浮选装置的喷射喷嘴上,或机械诱导浮选装置的转子上。

3. 流程配置

工艺配置是指工艺流程图,即储罐和容器的顺序、工艺流程的连接和路由,以及极为重要的废弃物流路由。对于一个给定的项目,可以设计出许多不同的工艺流程图。变化包括使用脱气容器、设计封闭排水系统、泵、阀门、提供动力流的仪表,以及对液位和流体流量的控制,这些都是工艺配置的一部分。在项目的设计阶段,工艺配置的开发称为工艺集成。通常,在确定设备对整个工艺的影响之前,不应选择任何设备。设计团队评估各种工艺流程图,最终选择最适合项目的工艺流程图 (Walsh)。

图 3.1显示了 GOM 各平台上系统组成的工艺流程图 (PFD) 。水力旋流器和浮选机用于油水分离。

图 3.1——复合工艺流程图,代表了墨西哥湾深水作业中通常安装的工艺系统的基本要素。
图 3.1——复合工艺流程图,代表了墨西哥湾深水作业中通常安装的工艺系统的基本要素。

在选择最合适的设施设计时,设计团队必须考虑一些从采出水中去除分散油的最佳实践。这些最佳实践包括:

  • 尽量减少阀门和泵中的液滴剪切。
  • 最大限度地延长剪切机构和下游分离设备之间的管道长度,以促进管道中液滴的聚结。
  • 对上游的流体加热以促进油水分离。
  • 如果可能的话,在系统早期分离水,以防止液滴剪切。
  • 防止固体产生并提供分离和去除固体的设备(例如锥形底部污泥池)。
  • 考虑在污水舱排放时使用正排量泵,而不是离心泵。
  • 提供有效的拒收处理系统。
  • 提供有效的化学处理系统。
  • 提供有效的监测和控制系统。

我们之所以关注废渣处理系统,只是因为我们发现许多现有设施在妥善处理废渣方面存在问题。从整个系统的角度来看,废渣是一种循环流。它代表了兼容性、污染、流速波动等方面的无数可能的过程问题。如果管理不当,它可能会导致整个系统在许多方面陷入困境,并导致出水水质不佳。

所有水处理设备都会产生废液流。在许多情况下,希望将废液流通过初级分离过程。在选择设备和设计油水分离系统时,重要的是要考虑所有废液流的流速和油水浓度。需要实现整个过程的材料和流量平衡。如果初级分离设备的尺寸不足以管理废液和初级生产流,则系统将出现瓶颈。在确定管理废液和初级生产流的容器尺寸时,必须使用实际的废液率值。如果容器尺寸不合适,操作员将被迫减少油保留时间、水保留时间或废液流速。这些步骤中的任何一个都会导致油和/或水的质量不佳。在深水海上平台上,通常只有三种排放流 - 干油管道、干气管道和用于向外排放的经过处理的采出水。

3.1显示了用于管理废弃物流的锥底污水池。污水池中的液体被循环回工艺流程。

正确设计污油箱以尽量减少工艺干扰非常重要。该油箱可用于提供沉淀时间和另一个化学处理机会。应提供加热流体至 120°F 或更高温度的装置,以改善油/水/固体的分离。

  • 应定期清除锥底罐中沉淀的固体并运往岸上进行最终处理。
  • 含有分离的油和油湿固体的化学稳定化乳化垫最好直接泵入输油管道或单独的储罐,以便在专门设计的滑轨中进一步进行海上或陆上加工。这不应再循环回工艺流中。通常,油性乳化液在泵入输油管道时会增加销售油的 BS&W,因此必须在油脱水规范的限制范围内进行。
  • 分离出来的水可以再循环回工艺流中,或再循环到专门设计的滑轨中的单独水箱中进一步处理。在以缓慢的速度(最好低于系统总流量的 5%)连续循环回工艺之前,应检查该流体流的水质,以尽量减少工艺干扰的可能性。来自污油箱顶部的澄清水应再循环到水处理系统的入口,而不是上游分离器,因为残留的化学品和固体可能导致界面乳液的形成,从而干扰处理工艺。

4. 操作输入

运营代表在所有阶段(从概念设计开始,到设计团队移交给运营部门)的积极参与至关重要。运营代表在设施运营方面拥有丰富的实践经验,能够就哪些措施有效、哪些措施需要改进提供非常有价值的意见(沃尔什)。他们能够提供出色的意见,尤其是在以下领域。

  • 从其他设施的设计和运营中吸取的经验教训。应将这些经验教训纳入改进设施设计中。
  • 详细的工程设计、设备选择和施工阶段。
  • 操作程序和实践的制定。
  • 调试和启动程序的开发。

运营的参与将极大地促进设施运营从设计团队向运营团队的移交。

5. 适合舷外处置的产出水水质

政府法规规定了船外处置所需的水质参数。例如,在美国联邦水域(法律上称为外大陆架 (OCS)),环境保护署 (EPA) 于 1972 年制定了《清洁水法案》(CWA),随后制定了 NPDES(国家污染物排放消除系统)许可计划。

墨西哥湾采出水的船外处置必须获得 NPDES 计划颁发的许可证授权,并且必须符合许可证中详述的监管准则。我们在下面简要介绍了这些准则。

  • 采出水排放中的油脂含量必须低于每日最高值 42 毫克/升和每月平均值 29 毫克/升。油脂监测样品应每月至少分析一次。
  • 受纳水体表面无可见光泽。应在观察到光泽后 2 小时内采集产出水样本并分析油和油脂。监管程序为 US EPA-1664。
  • 应根据《评估流出物和受纳水对海洋和河口生物的慢性毒性的短期方法》(EPA/821-R-02 014)或最新版本,通过 7 天慢性全流出物毒性 (WET) 测试来评估毒性。要符合 WET 限制,无可观察效应浓度 (NOEC) 必须等于或大于 NPDES 许可证附录 D 表 1 中规定的临界稀释(流出物百分比)浓度。

根据许可证表 1 计算出的临界稀释浓度取决于排放速率、排放管直径以及排放管与底部之间的水深。NOEC 定义为与对照样品相比不会导致致死的最大流出物稀释度。NPDES 许可证
详细说明了生产水采样方法、采样频率、样品保存以及样品分析的分析技术、毒性测试和报告要求。读者应参考 NPDES 许可证了解详情。

6. 结论

以下总结了第 1 部分和第 2 部分的主要结论。

1. 六个要素或主题是墨西哥湾海上平台采出水处理系统系统设计的基础。这些主题是流体特性、化学处理、设备、工艺配置、操作和流出物质量。

2. 流入流体的特性以及流出物的质量决定了水处理的挑战。

3. 在设计阶段,大多数情况下无法获得采出水化学数据、分散油和悬浮固体浓度以及粒度分布数据。

4. 必须通过对周围油藏的了解来确定模拟系统。应根据油藏工程、工艺工程和设计油田的流动保障信息调整模拟油田/油藏的采出水特性,以便了解采出水特性,从而达到设计目的。

5. 采出水的质量和产量在油田寿命期间会发生变化。系统设计应确保充分管理这些变化。

6. 应在分离设备各级的上游和下游提供高质量的取样点,以便在现场使用期间对设备进行有效的监测和故障排除。

7. 应按照最佳实践安装取样探头,以确保以环保的方式采集有代表性的样品。取样探头应易于取用和回收。

8. 应安装化学注入点来注入化学药品,并按照最佳实践关注产品/流出物质量、流量保证和资产完整性。

9. 注射点应配备注射管或喷嘴,以确保与流体流正确混合。这些应易于接近。注射管或喷嘴应易于在线取回。

10. 化学品选择应基于模拟生产水化学(用于系统设计)的实验室测试,最好模拟工艺设计的温度和压力。

11. 水力旋流器和浮选机通常用作深水海上作业中的初级和次级除油设备,非常适合生产水处理。

水力旋流器

  • 大多数海上水力旋流器装置都有一个包含多个衬管的压力容器。根据操作和维护方面的考虑,建议每个压力容器中的衬管数量应在 50 到 60 个之间,以处理总产水流量的约 25% 或最大 50,000 BWPD。压力容器的数量应为 n+1,其中n是处理总产水率所需的容器数量,额外的(备用)容器用于在容器因维护而停止运行时使用。
  • 应使用直径较小的衬管(例如 15 或 25 毫米)来有效去除大于 20 微米的油滴。
  • 建议水力旋流器入口的进料压力为 125-150 psig 或更高,以提供操作灵活性。
  • 一般认为约 2 的压差比和约 2.5 的拒收率接近最佳值。
  • 最好使用水力旋流器的最大额定流量的 80% 来控制由段塞效应引起的流体波动。

浮选机

  • 深水作业通常使用多级(四级)水平水力诱导浮选机和单室垂直诱导浮选机。水平浮选机中的流体停留时间约为 4 至 5 分钟。
  • 浮选装置可以有效去除大于 10 微米的油滴。
  • 浮选装置的效率直接取决于细胞停留时间和细胞数量。
  • 单室立式浮选装置的最大污染物去除率约为 70%,即使流体停留时间为 4 分钟也是如此。这应该是设计阶段的一个重要考虑因素。双室立式浮选装置更适合高效运行。
  • 四槽水平浮选机的污染物去除效率约为 90%。
  • 浮选装置的设计流量应包括用于将气体引入流体的循环流的流量。在多级水力诱导浮选槽中,循环流约占处理流体流的 25%。

12. 正确设计污油箱以处理废弃物流至关重要。设计应包括将分离出的固体拖运到岸上处理以及将乳化垫直接回收到销售油管线的规定,以尽量减少工艺干扰和脱油设备性能。13
. 采出水处理系统的设计应考虑污油箱的回收流率和回收液体的质量。

14. 运营代表应是设计团队的成员,设计团队应仔细考虑以前设施设计和运营中的最佳实践和经验教训,以改进设施设计。

15. 设计团队应以满足政府法规要求的污水处理产水水质为基础进行设施设计。

致谢

我们衷心感谢 Ken Arnold, PE,他花时间审阅我们的论文并提供专家建议和评论。这些非常有帮助。

Kris Bansal博士、SPE 是一名顾问,擅长开发最佳工程解决方案,这些解决方案需要系统集成并与上游作业中从油藏到顶部的多个学科进行交互。他在康菲石油公司担任工程研究员 30 年后退休,从事勘探与生产工作。他曾在 BP 担任顾问一年,为 BP 墨西哥湾作业提供采出水管理问题的技术支持。他还为 BP 的工程师提供采出水管理方面的培训。

在康菲石油公司,他是上游水资源管理领域的全球主题专家,为康菲石油公司的全球运营提供技术和解决问题的专业知识,以开发创新、经济高效且环保的工艺问题解决方案。他教授注水课程,并举办问题解决研讨会,将技术转移到康菲石油公司的运营和工程部门。

在加入康菲石油公司之前,他在沙特阿美公司(加瓦尔油田)从事运营工程工作 6 年,在卡尔冈公司从事活性炭研发工作 3 年。在进入工业界之前,他在卡内基梅隆大学和德国汉迈特纳核研究所从事学术研究工作 6 年。

他拥有物理化学硕士和博士学位以及化学工程硕士学位,并拥有丰富的运营经验,为组织提供独特的基础知识与第一手实践经验的结合。他的联系方式是krismbansal@yahoo.com

John Walsh博士、SPE,在水行业工作了近 30 年。他在壳牌工作了 20 多年,在加入壳牌之前曾在 Westvaco 造纸公司工作,后来又在 CETCo 能源服务公司工作,最近又加入了 Worley Consultants。在壳牌,他是上游水处理的全球主题专家。在担任该职位期间,他曾在数十个国家工作,负责监督这些国家的研发、故障排除和技术项目。他参与了各种水处理挑战,包括页岩、水底、提高采收率以及常规陆上和深水海上项目。

他有幸与杰出的生产水专家合作,这些专家给予他源源不断的鼓励,使他撰写了一本两卷的书,名为《生产水》。书的封面材料中列出了贡献者名单。

他曾担任生产水协会主席兼董事总经理。他曾任 SPE 董事会成员,并担任两门 SPE 水处理课程的指定讲师。他拥有约翰霍普金斯大学化学工程博士学位。

进一步阅读

掌握深水 GOM 的生产水管理——第 1 部分, 作者:JM Walsh 和 KM Bansal,JPT。

《采出水》,第 1 卷:基础知识、水化学、乳化液、化学处理,作者 JM Walsh。由 Petro Water Technology LLC 出版。

采出水,第 2 卷:设备、工艺配置、应用,作者 JM Walsh。由 Petro Water Technology LLC 出版。

《地面生产作业》,第 3作者:K. Arnold、M. Stewart。由 Gulf Professional Publishing 出版,Elsevier(2008 年)。

SPE 16642水力旋流器:一种生产水处理解决方案,作者:H. Meldrum。

OTC 5594水力旋流器:一种生产水处理解决方案,作者:H. Meldrum。

SPE 81135新一代气体浮选技术简介——案例研究,作者:S. Movafaghian、J. Chen、S. Wheeler 和 RW Guidry。

墨西哥湾外大陆架西部和中部新旧污染源及新排放者的 NPDES 通用许可证 (GMG290000),2023 年 5 月。

原文链接/JPT
Water management

Mastering Produced-Water Management in Deepwater GOM: 25 Years of Insights—Part 2

This article is the second of a two-part series on produced-water management in the Gulf of Mexico and covers four themes: equipment, process configuration, operations, and effluent quality.

Offshore oil rig in a large body of water
For many design decisions, the industry is lacking systematic and methodical approaches to treat and manage produced water.
HeliRy/Getty Images

This article is the second of a two-part series and covers four themes: equipment, process configuration, operations, and effluent quality.

Part 1 included an introduction, a brief description of six major themes for produced-water treatment system design, and a more detailed discussion of two themes: the characterization of feed produced water and chemical treatment.

This article includes Part 1’s abstract, and the Conclusion section of this article addresses both Parts 1 and 2.

Abstract

This two-part series provides guidelines for the systematic design and operation of produced- water systems specifically for deepwater Gulf of Mexico (GOM) platforms. Six elements or themes are highlighted: fluid characterization, chemical treatment, equipment, process configuration, operations, and effluent quality. The characteristics of the incoming produced water, together with the target effluent quality, define the water-treating challenge.

Due to the high cost of space and weight in the deepwater environment, water treatment must have high intensity and must be highly integrated with surrounding equipment. High intensity refers to equipment that can manage high volumetric flow rates, over short residence times, and which occupies a small footprint. The concept of highly integrated system design refers to system process design that integrates water-treatment equipment, process configuration, and chemical treatment into a high-performing single system that removes a very high fraction of the contaminants (mostly oil suspended in water).

Every opportunity is utilized to ensure that the produced water is treated to a high standard. Peak shaving, interface bleed, and break tanks that act as clarifiers are just a few of the opportunities that can be utilized to improve produced-water quality.

Equipment

Fundamental information is provided to assist the design team to select equipment for treating produced water to meet regulatory guidelines for overboard water disposal. This includes primary and secondary deoiling equipment—hydrocyclones and flotation cells (Walsh, Arnold, and Stewart).

Primary separation equipment (separators, hydrocyclones) depends on density difference, contaminant (droplet and particle) size, and fluid viscosity. Secondary (flotation) performance depends on contaminant size, interfacial adsorption (which depends on the contaminant chemistry and chemical treatment), bubble size, and total surface area of the bubbles (gas/water ratio). Tertiary separation equipment (such as filters, membranes, etc.) depends on contaminant (floc) size, media surface properties (which depends on the contaminant chemistry and chemical treatment), media surface area, superficial velocity, and details regarding backwash etc.

In the deepwater GOM, the design objectives are met largely through customized design, mainly of hydrocyclones and flotation. These technologies are considered to be “conventional” since they have been available for decades. However, detailed design features of these conventional technologies are not well-known and have only come to light through After Action Review and examination of best practices across the region.

For hydrocyclones, higher pressures (about 150 psig) at the inlet are required for greater flexibility in adjusting operating parameters such as pressure differential ratio and reject ratio for optimal liner efficiency.

Custom design features include small-diameter liners (for greater forward pressure drop and separation efficiency), staging of the throughput (0.25/0.5/ 0.75) has been used with success, and large reject port (for peak shaving). These features can be incorporated into the well-known conventional hydrocyclone.

For flotation, a wide range of approaches have been used to meet design objectives. Customization of flotation equipment includes methods for creating a large number of small-diameter bubbles; number of stages in the equipment design; injection/inline mixing of flotation chemicals; and valving and pipe flow design to minimize shearing. Through performance analysis, the design of single-stage flotation has been found to be inadequate for deepwater application.

The separation efficiency of tertiary equipment will not be discussed in this article because it is generally not required to meet overboard regulatory water-quality specifications and is not normally used in offshore operations.

1.   Hydrocyclones—Primary Deoiling Equipment

Hydrocyclones are regarded as “standard technology” in offshore operations for the removal of dispersed oil droplets from produced water and have been successfully used onshore as well. The oil-removal efficiency of a hydrocyclone is the difference between the oil-in-water (OiW) levels of the inlet stream and the water outlet stream divided by the OiW in the inlet stream (Walsh, Arnold, and Stewart).

1.1       Driving Force for Separation

Hydrocyclones utilize liners to generate high centrifugal forces (1,000–2,000 G, where G is the gravitational constant) to classify the influent oil/water stream by density.

Large centrifugal forces are generated by inducing a swirl motion to the fluid inside the liner. The angular acceleration due to the swirl motion enhances the effect of the density difference between the dispersed oil droplets and water. The oil, being lighter than water, migrates to the axial center of the liner. This core of oil flows in the opposite direction of the water effluent due to pressure difference between the feed and the reject orifice. As shown in Fig. 1.1, the oil is directed to a reject port (overflow), and the water exits through an effluent port (underflow in Fig. 1.1). The reject core is drawn out of the reject orifice at the top of the cyclone liner.

A modified Stokes’ Law equation for separation of oil droplets from water applies.

V0 = Fc * 1.78 * 10-6 (Dr) d2

Where Fc is the centrifugal force, multiples of gravity, V0 is the vertical velocity of the oil droplet in ft/sec, d is the droplet diameter in microns, ∆ρ is the specific gravity difference between water and oil, and µ is fluid viscosity in centipoise. The migration rate of the oil droplets to the central core (or axis) is controlled by the same variables in Stokes’ Law with one exception. Stokes’ Law is based on gravity as the driving force whereas in a hydrocyclone the centrifugal acceleration of the swirl motion provides the driving force.

Various Liner Sections

A liner has four sections—a cylindrical swirl chamber, a concentric reducing section, a fine tapered section, and a cylindrical tail section (Figure 1.1). The location between concentric reducing section and taper section is the point of cyclone size specification (hydrocyclone diameter). Fluid acceleration (tangential velocity) occurs in the reducing, taper, and parallel sections. The reducing and taper sections have a conical shape, and the parallel section has a cylindrical shape.

Schematic diagram of a single cyclone with labeled sections
Fig. 1.1—Schematic diagram of a single cyclone with labeled sections and showing geometry and fluid path of a typical deoiling cyclone.  Such a cyclone is typical of that used in packaged deoiling hydrocyclone units.

Packaging of Liners

Most offshore hydrocyclone installations have a pressure vessel that houses many liners (Fig. 1.2). The most common packaged unit has a single feed stream and a single product and reject stream.
 

Packaging of many cyclonic liners within a single pressure vessel. 
Fig. 1.2.—Packaging of many cyclonic liners within a single pressure vessel.  This cutaway diagram shows the reject header and the feed section below the reject header.  A single liner is shown in some detail, including the O-rings which seal the liner into the reject header.

Packaging is an important aspect of hydrocyclone installation. Manufacturers have developed many varieties of packaging which have various desirable features. The design team will need to review various packaging to select the most appropriate for the design project.

It is important to select a reasonable number of liners in each pressure vessel (such as between 50 and 60). Based on operational and maintenance considerations, it may be useful to select the number of liners in each pressure housing to process roughly 25% of the total produced-water rate or a maximum of about 50,000 BWPD.

It may be advantageous to provide a spare pressure vessel to be used when one of the vessels with liners is out of service for maintenance, especially in offshore operations.

1.2    Operating Principles

1.2.1 Separation, Capacity, and Equipment Cost

Separation efficiency depends strongly on the diameter of the liner. The smaller the diameter, the lower the liner throughput capacity, but the higher the swirl velocity and the greater the driving force for oil/water separation. Lower capacity means more liners will be required for a given water flow rate, and subsequently, equipment cost will be higher.

In deepwater operations, droplet size is very small because of excessive shearing and limited room for long pipe runs to allow droplet coalescence. Based on this, smaller-diameter (high-efficiency) liners are preferred. That is one of the basic decisions that will need to be made when selecting the liners.

The curves in Fig. 1.3 were generated for different liner sizes based on empirical correlations for a specific set of density difference and water viscosity. Differences in these variables will change the results but not the overall conclusions.

Fig. 1.3—Typical separation efficiency curves for various hydrocyclone designs. 
Fig. 1.3—Typical separation efficiency curves for various hydrocyclone designs. 

For deepwater offshore operations, smaller-diameter hydrocyclones (such as 10 mm or 25 mm) are preferred to remove small-diameter (<20 microns) dispersed oil droplets.

1.2.2 Flow Rate and Hydrocyclone Separation Efficiency (Operating Envelope)

Turndown Ratio. As shown in Fig. 1.4, the separation efficiency of a hydrocyclone typically increases as the flow rate increases and remains constant over a wide range of flow rates. At a certain high flow rate, the separation efficiency starts to drop rapidly. The hydrocyclone designs from each manufacturer have their unique curve. The curves in Fig. 1.4 show a turndown ratio which is approximately 4 to 1. For operational purposes, it is best to use the hydrocyclone at 80% of its maximum rated flow rate to handle fluid rate fluctuations from slugging generally observed in operations.

Fig. 1.4—The separation efficiency as a function of the flow rate over a wide range of flow rates.
Fig. 1.4—The separation efficiency as a function of the flow rate over a wide range of flow rates.

Pressures

Three pressures are important to the optimization of fluid flow and separation efficiency in the hydrocyclone: (a) forward or inlet pressure, (b) treated effluent or underflow pressure, and (c) reject or overflow pressure.

Underflow pressure is a function of the inlet pressure and volumetric flow rate of liquid through the hydrocyclone, which creates a pressure drop.

In a typical installation, the hydrocyclone is connected to a separation vessel with the level-control valve of that vessel downstream of the hydrocyclone. Thus, the inlet pressure is the pressure in the vessel and the underflow pressure becomes the upstream pressure to be used in sizing the vessel’s level-control valve.

The difference between the inlet pressure and the underflow pressure is called the forward pressure drop. The importance of the forward pressure drop cannot be overstated yet it is often specified incorrectly. This pressure drop should be maximized since it provides the driving force for separation. The inlet pressure is provided by an upstream separator or a suitably (low-shear) selected pump and is the highest of the three pressures. For design purposes, it is preferred that this feed pressure is greater than about 125 psig or preferably around 150 psig. High inlet pressure pushes the fluid through the liner, generating high centrifugal forces.

Low pressure in the fluid core permits the reverse flow of the oily water to the reject port (reject orifice). The reject port is a very small orifice that is about 1–3 mm in diameter. This is designed to ensure that the hydrocyclone reject rate is small and only the fluid from the central oily core flows in the direction of the overflow.

Overflow pressure is controlled by external valves and the valve-control system.

Reject Ratio. The ratio of the reject or overflow rate to the inlet flowrate is called the reject ratio. It is modulated by adjusting valves on the overflow and underflow discharge piping.

Reject Ratio = (Overflow rate/Inlet flow rate) * 100

Hydrocyclone systems have an optimum reject ratio. Systems operating at less than the optimum reject ratio will result in lower oil removal efficiency. A margin of safety may be determined by quantifying the decline in removal efficiency as a function of the reject ratio. In most cases, this reject ratio is about 2.5.

Dispersed particles in the oily reject stream can result in plugging of small reject orifices (about 2 mm) and impact oil-removal efficiency. This can be corrected by periodic backflushing of the reject orifices.

Pressure Differential Ratio (PDR)

PDR = (Pinlet – Preject) / (Pinlet – Pwater outlet)

Pinlet is the inlet pressure, Preject is the pressure at the reject port or Poverflow, Pwater outlet   is the pressure at the water outlet stream or P underflow. For oil and water separation, the optimal PDR is between 1.7 and 2.0, and the exact PDR will need to be determined during operation.

PDR must be greater than 1 to force the oil in the core to go in the reverse direction of flow. For a given hydrocyclone geometry, the PDR controls the amount of fluid that is rejected. Fig. 1.5 provides insight into the relationship between the PDR, reject flow rate, and the separation efficiency. The figure on the right is the relation between the reject ratio and the PDR. Fig. 1.5 shows the relationships between two particular hydrocyclone designs.

Figure 1.5—Two diagrams together demonstrate the effect between DP ratio (PDR) and separation efficiency. The two of these variables are tied together through the reject ratio.
Figure 1.5—Two diagrams together demonstrate the effect between DP ratio (PDR) and separation efficiency. The two of these variables are tied together through the reject ratio.

Performance Data. Field data for hydrocyclone performance on the Auger platform in 2004 is presented in Fig. 1.6. The data shows that the primary benefit of the hydrocyclone is to significantly reduce the oil concentration downstream of it (peak shaving). This can significantly improve the performance of the downstream water-treatment equipment such as flotation cells. The gradual deterioration of hydrocyclone performance (squares) that started in October 2004 was due to a reduced backwash frequency and lower flow rates without adjustment of the number of liners.

Field data. Oil concentration as measured by IR at the FWKO discharge
Fig. 1.6—Field data. Oil concentration as measured by IR at the FWKO discharge (diamonds, hydrocyclone feed) and the hydrocyclone discharge/flotation feed (squares).  Note that FWKO discharge (feed to hydrocyclone) attains very high values. The main benefit of the hydrocyclone is to “peak shave.” This improves flotation performance.

1.3. Hydrocyclones—Advantages and Drawbacks

1.3.1 Advantages

  • Peak shaving
  • Fluid residence time about 2 sec
  • Flexible modular design.
  • Easy to operate and maintain. No moving parts.
  • Removes oil droplets >15 microns very efficiently in most cases
  • Broad and predictable operating range
  • Insensitive to platform motion
  • Insensitive to slugging
  • Quick startup and recovery from upsets
  • Oily water reject rate is about 2%

1.3.2 Disadvantages

  • Sensitivity of the hydrocyclone system efficiency on droplet size distribution
  • Requires relatively high pressure (about 150 psig) for optimal hydrocyclone performance
  • Requires a low-shear pump to boost operating pressure in low-pressure systems
  • Small-diameter reject port prone to plugging, thus requires regular backflushing to minimize blockage potential
  • Performance deteriorates in the presence of oil wet solids

2.   Flotation Cells—Secondary Deoiling Equipment

In this section, we briefly describe the basic information that the design engineer should consider for selection of flotation equipment for a given application. We do not describe details of various types of flotation equipment but simply present guidelines that will assist design engineers in selecting equipment.

Flotation is a secondary process. If hydrocyclones are present in the system, flotation is always located downstream of the hydrocyclones.

Flotation is typically capable of separating small amounts of oily solids along with dispersed oil particles and thereby increases a system’s efficiency. Gas flotation equipment is generally more efficient in dispersed oil removal when compared with liquid-liquid hydrocyclones because of its ability to remove both dispersed oil droplets and at least some oily solids.

Flotation units utilize finely dispersed gas bubbles to carry dispersed oil droplets and oily solids to the surface. Oil droplets and oily solids adhere to the surface of the bubbles and rise to the surface with a velocity that is estimated from Stokes’ Law. The attachment of dispersed gas bubbles to the oil droplet and oily solids reduces the specific gravity, thereby producing a faster rise and accelerating their separation.

Chemicals are used to coagulate and flocculate the dispersed droplets. Gas bubbles and flocculated droplets form a froth layer on the water surface, and contaminants on the surface are removed by skimming.

2.1       Mechanism of Separation

In a flotation cell, gas bubbles are injected or induced. Gas bubbles rise rapidly and collide with the oil drops. The collision frequency depends on the concentration of oil drops, the concentration of gas bubbles, and on the projected areas of the oil drops and the gas bubbles. Very few collisions lead to the capture of oil drops by gas bubbles. This is known as the “capture efficiency.” This is roughly 1/1000. It is important to note that capture efficiency is much smaller when large bubbles are used when compared with the capture efficiency when small bubbles are used. Capture efficiency depends on the surface chemistry of oil and gas. It is strongly affected by chemical additives such as coagulants and flocculants.

Fig. 2.1—Schematic illustration of gas bubbles rising (blue dots) and oil droplets. The collision path of two of the bubbles is shown.
Fig. 2.1—Schematic illustration of gas bubbles rising (blue dots) and oil droplets. The collision path of two of the bubbles is shown.
Fig. 2.2—Collision frequency and capture efficiency.
Fig. 2.2—Collision frequency and capture efficiency.

Fig. 2.2 is a schematic showing the impact of gas-to-water ratio and bubble size on collision frequency and capture efficiency. For improved separation of dispersed oil droplets from produced water in a flotation cell, small bubbles and high gas/water/oil ratio are needed.

2.2       Design Considerations

Flotation is a contact separation process. It relies on the contact of gas bubbles with oil drops and solid particles. The important variables are:

  • Gas/water ratio (volumetric ratio)
  • Bubble size (diameter)
  • Staging (single, dual, multistage)
  • Chemical application (surface chemistry), such as coagulants and flocculants
  • Contaminant size and surface chemistry

Design variables include large contaminant oil droplets and oily solid particles, uniform bubble distribution throughout the flotation cell, high gas/water ratio, and small (optimal size) gas bubbles. The bubble size should be minimized while ensuring that the bubbles are large enough to migrate to the oil/water interface. The contaminant size can be increased via chemical usage such as coagulation and flocculation. These chemicals are injected downstream of the hydrocyclones but upstream of the flotation cells. The use of multiple stages in a flotation cell is a very effective way of increasing the separation efficiency. For example, in most applications, horizontal hydraulically and mechanically induced flotation cells routinely use four separation stages. The total fluid residence time in a flotation cell is about 5 minutes.
In addition to the aforementioned variables, there are a few important practical considerations in the mechanical design.

  • The oily water must be introduced in a way that does not disrupt the flotation process.
  • The wet oil and clean effluent water must be discharged in a way that does not allow contamination with the feed and that does not disrupt the process.
  • The overall weight and space must be as low as possible, while achieving a high separation performance.
  • The reject stream from the flotation cell must be properly handled (such as in a properly designed slop tank) to ensure minimal impact on the overall produced-water treatment system. This reject stream in a multistage flotation unit is normally between 5 and 10% of the incoming flow rate to the cell.

As with most water-treatment equipment, integration of the flotation unit into a facility or process is critical to successful performance. The most important aspect of integration is handling the reject.

2.3       Flotation Equipment

There are four types of gas flotation cells that are used in the upstream oil and gas industry (Walsh Arnold, and Stewart).

  • Dissolved gas flotation
  • Horizontal multistage hydraulically induced flotation
  • Horizontal multistage mechanically induced flotation
  • Vertical induced gas flotation—commonly used offshore in deepwater operations

Almost all horizontal units are multistage. We do not provide any details on the equipment but will simply highlight some features that can help design engineers in selecting the most appropriate unit for the application.

2.3.1    Dissolved Gas Flotation (DGF) Unit

In a DGF, gas is dissolved into a recirculating recycle clean-water stream (about 20– 25%) at high pressure (around 20–40 psig). Natural gas is generally used to exclude oxygen. The gas- saturated water then passes through a valve and into the flotation chamber. The dissolved gas breaks out of the solution in small-diameter bubbles (20 to 60 microns) when the flow enters the chamber. These bubbles are much smaller than in a typical induced-gas device which are approximately five to 10 times larger. Dissolved gas generally is between 0.2 to 0.5 scf/bbl of water to be treated.

The amount of gas that can be injected into a DGF is limited due to two factors. The first is a thermodynamic limitation of gas solubility that depends on the temperature of the water and pressure of gas. The second factor is the practical limit in how much treated water can be recycled to inject the gas into the feedwater. Although small bubbles are generated in the DGF, the total number of bubbles is relatively small because of the limitations of volume of the injected gas. So, in most cases, the overall separation efficiency is around 50%. DGFs can be single or multistage units.

DGFs are not commonly used the upstream oil and gas industry because of low separation efficiency.

2.3.2    Horizontal Hydraulically Induced Flotation Unit

In these units, gas dispersion is achieved by circulating a portion (about 25%) of the clean effluent water in each cell section. This water re-enters the flotation cell through a venturi which draws gas from the vapor space over the water in the cell and discharges it with the water into the bottom of the cell. These hydraulic units are mostly four cells.

The capacity of the unit must be designed to handle the feed (untreated influent) plus the recirculated water.

2.3.3    Horizontal Mechanically Induced Flotation Units (Such as Wemco)

These units achieve gas dispersion with a rotor turned by an electric motor. As the rotor turns, it acts as a pump forcing water through a disperser and creating a vacuum which pulls gas from the gas blanket into a standpipe and mixes it with water. As the gas-water mixture moves through the disperser, small gas bubbles form. The gas bubbles thus formed gather the oily floc and bring it to the surface where it is skimmed off. In general, the efficiency of a mechanically induced gas flotation cell is higher than the hydraulically induced cell. However, due to higher maintenance costs associated with multiple motors and rotors, hydraulic units are often preferred over mechanical units.

2.3.4 Staging and Removal Efficiency

As mentioned before, horizontal flotation units are generally multistage. In a four-cell unit, the total fluid retention time is typically 5 minutes. Bulk water flow moves in series from one cell to the other by underflow baffles. Water flow from one cell to the other is basically in a plug flow mode.

Flotation kinetic analysis has long been used to predict the contaminant removal efficiency in an IGF (induced gas flotation) process. Based on this analysis, efficiency is a direct function of the cell residence time and the number of cells (Movafagian et al). Fig. 2.3 is based on the kinetic analysis. It shows that in a single cell unit, the maximum contaminant removal efficiency is about 70%, even if the fluid residence time is increased to 2.5 minutes. The design engineer should take this into consideration.

Fig. 2.3—Contaminant removal efficiency as a function of cell residence time and number of cells.
Fig. 2.3—Contaminant removal efficiency as a function of cell residence time and number of cells.

2.3.4 Vertical Induced Gas Flotation Units (VIGF)

VIGFs (typically single stage) were originally designed to address concerns about the effect of wave motion on deepwater floating structures, to reduce footprint, and as a lower-cost alternative to the large multistage horizontal units discussed above. VIGFs have performed well; however, single-vessel applications can be prone to upsets, offering no redundancy or protection against mechanical failure.

When vertical flotation units are required to meet system design or space limitations, the use of two or more vessels in series should be considered because of low efficiency for single-cell units as shown in Fig. 2.3. Currently, the main advantages offered by vertical IGF are low cost and reduced space requirement and weight.

Gas bubbles in a typical VIGF are generated either by using eductors (hydraulically induced such as in a Unicell), or in some designs, specialized pumps are used to generate gas bubbles. However, in all these VIGFs, single stage is used and the fluid residence time varies between 2 to 5 minutes.

The vertical configuration lends itself to the use of vessel internal elements that promote uniform distribution of gas, good gas bubble/oil drop contact, and coalescence of oil drops. This has led to sophisticated internal elements, which make these units significantly different from the classic horizontal IGF units. A major disadvantage of using coalescing elements is their tendency to become plugged by suspended solids present in produced water.

2.4       Flotation Cells—General Performance Guidelines

  • The efficiency of a four-unit flotation cell is higher than a single unit cell.
  • Induced gas flotation units have higher efficiency than dissolved gas units.
  • Experience has shown that mechanically induced flotation units have higher efficiency than hydraulically induced flotation units.
  • In general, oil droplets greater than about 10 microns are efficiently removed in flotation units.
  • A proper design and operation of induced gas flotation units should be capable of around 90% dispersed oil removal.

2.5       Flotation Cell Advantages and Disadvantages

Advantages

  • Well-known process with long operating history
  • Function does not depend entirely on the density difference between the oil and the water phase
  • Good turndown (longer residence time, improved separation)
  • Effective system for removing dispersed oil, especially when used with chemical addition (coagulants and flocculants)
  • Generally, achieves, if properly operated and maintained and used in conjunction with properly designed primary deoiling equipment, a dispersed oil concentration in overboard water of less than 29 mg/L
  • Oil droplets >about 10 microns are efficiently removed

Disadvantages

  • Fluid residence time is relatively long (around 5 minutes). Equipment is relatively heavy and bulky.
  • Not suited for floating production systems.
  • Needs a steady constant feed. Vulnerable to upset shock loads of oil with subsequent slow recovery.
  • Vulnerable to changes in flow and water composition which may result from startup/shutdown of individual wells.
  • Carryover of oily floc from flotation cells is difficult to eliminate, which can degrade effluent water quality and can cause sheen on the surface of receiving waters.
  • Generally, very sensitive to chemicals and their dosages.
  • Manpower-intensive with respect to operation and maintenance.
  • Shearing of oil droplets may occur across eductor nozzle in hydraulically induced cells or rotor in mechanically induced flotation units.

3.   Process Configuration

Process configuration refers to the process flow diagram, i.e., the sequence of tanks and vessels, connections and routing of process flows, and the all-important routing of reject streams. Many different process flow diagrams could be devised for a given project. Variations include use of degassing vessels, design of closed drain system, pumps, valves, instrumentation that provides motive flow, and control of levels and fluid flow are all part of the process configuration. In the design stage of a project, the development of a process configuration is known as process integration. Generally, no piece of equipment should be selected until the impact of that equipment on the overall process has been determined. The design team evaluates various process flow diagrams and finally selects the one that is most appropriate for the project (Walsh).

A process flow diagram (PFD) which is a composite of systems on various GOM platforms is shown in Fig. 3.1. Hydrocyclones and flotation cells are used for oil/water separation.

Fig. 3.1—A composite process flow diagram representing the essentials of the process system typically installed in GOM deepwater operations.
Fig. 3.1—A composite process flow diagram representing the essentials of the process system typically installed in GOM deepwater operations.

It is important for the design team to take into consideration some of the best practices for removing dispersed oil from produced water during selection of the most appropriate design for the facility. These best practices include:

  • Minimize droplet shear in valves and pumps.
  • Maximize pipe lengths between shearing mechanisms and downstream separation equipment to promote pipe coalescence of droplets.
  • Apply heat to the fluids far upstream to facilitate oil/water separation.
  • Separate water early in the system, if possible, to prevent droplet shearing.
  • Prevent solids production and provide equipment to separate and remove solids (e.g., conical bottom slop tank).
  • Consider the use of positive displacement pump rather than a centrifugal pump on the discharge of the slop tank.
  • Provide an effective rejects-handling system.
  • Provide an effective chemical-treatment system.
  • Provide an effective monitoring and control system.

We focus on the rejects-handling system only because of issues that we have observed in properly handling the rejects in many existing facilities. From the standpoint of overall system, the reject is a recycle stream. It represents an endless list of possible process problems in terms of compatibility, contamination, flow-rate fluctuations, and so on. If not managed properly, it can cause the entire system to struggle in many ways and cause poor effluent water quality.

All water-treating equipment generates a reject stream. In many cases, it is desirable to route the reject stream through the primary separation process. In selecting equipment and designing an oil/water separation system, it is important to consider the flow rate and oil-in-water concentration of all reject streams. A material and flow balance of the overall process is required. If the primary separation equipment is not adequately sized to manage both the reject and the primary production stream, then the system will be bottlenecked. Realistic values of reject rate must be used in sizing the vessels which will manage both the reject and the primary production stream. If the vessels are not properly sized, the operators will be forced to reduce either the oil-retention time, the water-retention time, or the reject flow rate. Any one of these steps will result in poor oil and/or water quality. On a deepwater offshore platform, there are typically only three discharge streams—dry-oil pipeline, dry-gas pipeline, and treated produced water for overboard discharge.

In Fig. 3.1, a cone-bottom slop tank is shown for managing the reject stream. Fluids from the slop tank are recycled back into the process stream.

It is important to properly design the slop tank to minimize process upsets. This tank can be used to provide settling time and another opportunity for chemical treating. Provisions should be provided for heating the fluids to a temperature of 120°F or higher for improved separation into oil/water/solids.

  • The settled solids in the cone bottom tank should be periodically removed and hauled for eventual disposal onshore.
  • The chemically stabilized emulsion pad containing separated oil and oil wet solids should preferably be pumped directly into the oil pipeline or to a separate tank for further processing offshore or onshore in a specially designed skid. This should not be recycled back into the process stream. Typically, the oily emulsion when pumped into the oil pipeline will contribute to the BS&W of the sales oil, so this must be done within the limits of the oil dehydration specification.
  • The separated water can be recycled back into the process stream or to a separate tank for further processing in a specially designed skid. Water quality for this fluid stream should be checked before it is recycled back into the process continuously at a slow rate (preferably less than 5% of the total rate through the system) to minimize the potential for process upsets. Clarified water from the top of the slop tank should be recycled to the inlet of the water-treatment system and not to an upstream separator where residual chemicals and solids can contribute to the formation of the interface emulsions which can upset the treatment process.

4.   Operations Input

It is extremely important that the design team has very good participation from the operations representatives in all phases, starting with the conceptual design and ending when the design team hands over to operations. Operations representatives have enormous practical experience in operating the facilities and can provide very valuable input on what works and what needs improvement (Walsh). They can provide excellent input, especially in the following areas.

  • Lessons learned from the design and operations of other facilities. These should be incorporated for improved facility design.
  • Detailed engineering design, equipment selection, and construction phase.
  • Development of operating procedures and practices.
  • Development of commissioning and startup procedures.

Operations participation will greatly facilitate the handover of the facility operation from the design team to operations.

5.   Effluent Produced-Water Quality for Overboard Disposal

Governmental regulations govern the water-quality parameters needed for overboard disposal. For example, in the US federal waters (designated legally as the Outer Continental Shelf (OCS)), the Environmental Protection Agency (EPA) created the Clean Water Act (CWA) in 1972, which then created the NPDES (The National Pollutant Discharge Elimination System) permitting program.

Overboard disposal of produced water in the GOM must be authorized by a permit issued under the NPDES program and must meet regulatory guidelines detailed in the permit. We briefly describe these guidelines below.

  • The oil and grease content of produced-water discharge must be below a daily maximum of 42 mg/L and a monthly average of 29 mg/L. Samples for oil and grease monitoring shall be analyzed at least once per month.
  • No visible sheen on the surface of the receiving waters. A produced-water sample shall be collected, within 2 hours of when a sheen is observed and analyzed for oil and grease. The regulatory procedure is US EPA-1664.
  • Toxicity shall be assessed through a 7-day chronic Whole Effluent Toxicity (WET) test in accordance with Short Term Methods for Estimating the Chronic Toxicity of Effluents and Receiving Water to Marine and Estuarine Organisms (EPA/821-R-02 014), or the most current edition. To be in compliance with a WET limit, the No Observable Effect Concentration (NOEC) must be equal to or greater than the critical dilution (percent effluent) concentration specified in Appendix D, Table 1 of the NPDES permit.

The critical dilution concentration calculated from Table 1 of the permit depends on the discharge rate, discharge pipe diameter, and the water depth between the discharge pipe and the bottom. The NOEC is defined as the greatest effluent dilution which does not result in lethality, compared to the control sample.
Produced-water sampling methods, sampling frequency, sample preservation, and analytical techniques for sample analysis, toxicity testing, and reporting requirements are detailed in the NPDES permit. The reader should refer to the NPDES permit for details.

6.    Conclusions

Major conclusions from Part 1 and Part 2 are summarized below.

1. Six elements or themes are the basis for the systematic design of produced-water-treatment systems for offshore GOM platforms. These themes are fluid characterization, chemical treatment, equipment, process configuration, operations, and effluent quality.

2. The characteristics of the incoming fluids, together with the effluent quality, define the water-treating challenge.

3. In most cases during the design phase, data on produced-water chemistry, dispersed oil and suspended solids concentration, and particle size distribution data are not available.

4. Analog systems must be identified through knowledge of surrounding reservoirs. Produced-water characteristics from analog fields/reservoirs should be adjusted based on the information from reservoir engineering, process engineering, and flow assurance from the field under design for developing an understanding of produced-water characterization for design purposes.

5. Produced-water quality and volumes vary during the field life. The system design should ensure that these variations are adequately managed.

6. Good-quality sampling points both upstream and downstream of all stages of separation equipment should be provided for effective monitoring and troubleshooting of equipment during field life.

7. Sampling probes should be installed in accordance with best practices to ensure representative samples can be taken in an environmentally acceptable manner. They should be readily accessible and easily retrievable.

8. Chemical-injection points should be installed to inject chemicals with a focus on product/effluent quality, flow assurance, and asset integrity in accordance with best practices.

9. Injection points should have injection quills or nozzles to ensure proper mixing with the fluid stream. These should be readily accessible. The injection quills or nozzles should be easily retrievable on-line.

10. Chemical selection should be based on laboratory testing with simulated produced-water chemistry (used for system design) and should preferably be done simulating process-design temperature and pressure.

11. Hydrocyclones and flotation cells, commonly used as primary and secondary deoiling equipment in deepwater offshore operations, are well suited for produced-water treating.

Hydrocyclones

  • Most offshore hydrocyclone installations have a pressure vessel containing several liners. Based on operational and maintenance considerations, suggested number of liners should be in the range of 50 to 60 in each pressure vessel to process about 25% of the total flow rate of produced water or a maximum of 50,000 BWPD. The number of pressure vessels should be n+1 where n is the number of vessels required to treat the total produced-water rate, and the extra (spare) vessel is for use when a vessel is off-stream for maintenance.
  • Smaller-diameter liners (such as 15 or 25 mm) should be used to efficiently remove oil droplets greater than 20 microns.
  • A feed pressure of 125–150 psig or higher at the inlet of the hydrocyclone is recommended to provide operations flexibility.
  • A pressure differential ratio of about 2 and a reject ratio of about 2.5 are generally considered to be near optimal.
  • It is best to use hydrocyclones at 80% of their maximum rated flow rate to manage fluid fluctuations resulting from slugging.

Flotation Cells

  • Multistage (four stages) horizontal hydraulically induced and single-cell vertical induced flotation units are commonly used in deepwater operations. Fluid residence time in a horizontal flotation unit is about 4 to 5 minutes.
  • Oil droplets greater than about 10 microns are efficiently removed in a flotation unit.
  • The efficiency of a flotation unit is a direct function of cell residence time and the number of cells.
  • Maximum contaminant removal in a single-cell vertical flotation unit is about 70%, even at a fluid residence time of 4 minutes. This should be an important consideration during the design phase. A two-cell vertical flotation unit is more suited for efficient operation.
  • A four-cell horizontal flotation cell has an efficiency of about 90% for contaminant removal.
  • The design rate of the flotation unit should include the rate of the recycle stream used for inducing gas into fluid. In a multistage hydraulically induced floatation cell, the recycle stream is about 25% of the treated fluid stream.

12. It is critical to properly design a slop tank to handle the rejects stream. Design should include provisions for hauling of the separated solids for onshore disposal and for recycling of the emulsion pad directly to the sales oil line to minimize process upsets and deoiling-equipment performance.
13. Design of a produced-water-treatment system should consider the rate of the recycle streams from the slop tank and the quality of the fluids being recycled.

14. Operations representative should be a member of the design team, and the input on best practices and lessons learned from design and operation of previous facilities should be carefully considered by the design team for improved facility design.

15. Design teams should base their facility design on meeting the effluent treated produced-water quality as required by governmental regulations.

Acknowledgment

Our sincere thanks to Ken Arnold, P.E., for taking time to review our paper and for providing expert advice and comments. These were very helpful.

Kris Bansal, PhD, SPE, is a consultant with expertise in developing optimum engineering solutions requiring system integration and interface with multiple disciplines from reservoir to the topsides in upstream operations. He retired as an Engineering Fellow from ConocoPhillips after 30 years of service in E&P. He spent a year at BP as a consultant providing technical support on produced water management issues in support of BP’s Gulf of Mexico operations. He also provided training in produced water management to BP’s engineers.

At ConocoPhillips, he was the global subject matter expert in upstream water management, providing technical and problem-solving expertise in ConocoPhillips worldwide operations to develop innovative, cost-effective, and environmentally sound solutions to process problems. He taught waterflood school and presented problem solving seminars for technology transfer to operations and engineering in ConocoPhillips.

Before joining ConocoPhillips, he spent 6 years in Saudi Aramco (Ghawar Field) in operations engineering and 3 years at Calgon Corporation in activated carbon R&D. Prior to joining industry, he spent 6 years in academic research at Carnegie-Mellon University and Han Meitner Institute of Nuclear Research in Germany.

Having earned an MS and PhD in physical chemistry and MS in chemical engineering, together with extensive experience in operations, he provides a unique combination of fundamental knowledge with firsthand practical experience to organizations. He can be reached at krismbansal@yahoo.com.

John Walsh, PhD, SPE, has worked in the water industry for close to 30 years. He worked for Shell for more than 20 years, Westvaco Paper Company prior to Shell, CETCo Energy Services, and most recently Worley Consultants. At Shell, he was the global subject-matter expert for upstream water treatment. In that role, he worked in dozens of countries, providing oversight for their R&D, troubleshooting and technical programs. He was involved in a wide range of water-treatment challenges including shale, water floor, enhanced oil recovery, and conventional onshore and deepwater offshore projects.

He had the great fortune to work with outstanding produced-water specialists who provided a steady stream of encouragement to write a two-volume book titled “Produced Water.” A list of contributors is given in the cover material of the book.

He was the president and managing director of the Produced Water Society. He has served on the SPE Board of Directors and is the designated instructor for two SPE courses on water treatment. He earned a PhD in chemical engineering from the Johns Hopkins University.

For Further Reading

Mastering Produced Water Management in Deepwater GOM—Part 1 by J.M. Walsh and K.M. Bansal, JPT.

Produced Water, Vol. 1: Fundamentals, Water Chemistry, Emulsions, Chemical Treatment by J.M. Walsh. Published by Petro Water Technology LLC.

Produced Water, Vol. 2: Equipment, Process Configuration, Application by J.M. Walsh. Published by Petro Water Technology LLC.

Surface Production Operations, 3rd Edition, by K. Arnold, M. Stewart. Published by Gulf Professional Publishing, Elsevier (2008).

SPE 16642 Hydrocyclones: A Solution to Produced Water Treatment by H. Meldrum.

OTC 5594 Hydrocyclones: A Solution to Produced Water Treatment by H. Meldrum.

SPE 81135 Introduction of a New Generation of Gas Flotation—A Case Study by S. Movafaghian, J. Chen, S. Wheeler, and R.W. Guidry.

The NPDES General Permit for New and Existing Sources and New Dischargers for the Western and Central Portion of the Outer Continental Shelf of the Gulf of Mexico (GMG290000), May 2023.