Performance-based quantification of natural draught wet cooling tower fill fouling

Date
2020-12
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH ABSTRACT: Natural draught wet cooling tower (NDWCT) fill fouling is a widely documented but poorly quantified phenomenon in the power generation industry. Specifically, there are no known methods enabling fair comparison of the fouling behaviour of competing fill designs. Using Monte Carlo methods, it is demonstrated that the thermal (Merkel number) and hydraulic (pressure loss factor) effects of fill fouling may be quantified separately, which introduces generally small errors in NDWCT performance prediction calculations. Fill fouling is subsequently characterised by individual time-dependent thermal and hydraulic fouling factors applied to the fill performance characteristic equations used in NDWCT performance prediction calculations. Fouling factors may be calculated, as in this study, using fill performance measured during long-duration fill testing, or may be estimated using predictive methods. It is demonstrated how the combination of predicted fouling factors and power plant parameters like heat rate correction and fuel cost may be used to perform fill lifecycle cost calculations, enabling fair fill comparisons. To provide confidence in the measured performance of NDCWT fills, a general method of fill performance measurement uncertainty analysis is developed. Earlier work has calculated the uncertainty of fill performance parameters at a single test operating point in a fill performance test. The current study expands significantly on this method, calculating the uncertainty at all test operating points and propagating these uncertainties first to the uncertainty of the fill thermal and hydraulic performance characteristic equations (known as regression uncertainty) and ultimately to the predicted NDWCT performance uncertainty. It is determined that the inclusion of correlated uncertainties is vital for the regression uncertainty calculations, but that they may be neglected at the NDWCT performance uncertainty calculation stage with relatively small errors. The developed uncertainty analysis method provides rigorous justification for performance-related commercial decisions based on fill performance testing. The developed fouling factor and uncertainty analysis methods are applied to the weekly performance measurements of four fill samples tested for a period of approximately 40 weeks at a fill fouling test facility located at a coal-fired power station in South Africa. This facility exposes the fill samples to operating conditions closely resembling true fill operation, including the use of actual power plant cooling water. While there are limitations to the testing data due to the relatively short duration and the low-fouling nature of the installed fill samples, as well as testing problems resulting in largely unusable thermal testing data, the use of the developed methods are thoroughly demonstrated.
AFRIKAANSE OPSOMMING: Natuurlike trek nat koeltoring (NTNKT) pakkingsaangroei is ‘n wydverspreide, maar swak gekwantifiseerde verskynsel in die kragopwekkingsindustrie. Daar is geen bekende metodes om die aangroei van kompeterende pakkings-ontwerpe regverdig te vergelyk nie. Dit word gedemonstreer, deur middel van Monte Carlo-metodes, dat die termiese (Merkel nommer) en hidrouliese (drukverliesfaktor) effekte van pakkingsaangroei apart gekwantifiseer kan word, wat oor die algemeen tot klein foute in die NTNKT verrigtingsvoorspellingsberekeninge lei. Pakkingsaangroei word gevolglik gekarakteriseer as individuele, tydafhanklike termiese en hidrouliese aangroei-faktore wat op die karakteristieke pakkingsverrigtingsvergelykings in die NTNKT verrigtingsvoorspellingsberekeninge toegepas word. Aangroei-faktore word, soos in hierdie studie, vanuit die gemete data van langtermyn pakkingstoetse bereken, of word beraam deur voorspellingsmetodes. Daar word gedemonstreer hoe die kombinasie van voorspelde pakkingsaangroeifaktore en kragstasie-parameters soos die warmtetempo korreksie en brandstofkoste gebruik kan word om die lewenssikluskoste van pakkings te bereken, wat die regverdige vergelyking van pakkingsontwerpe moontlik maak. Om vertroue in die gemete verrigting van NTNKT pakkings te verbeter, word ‘n algemene metode vir die onsekerheidsanalise van pakkingsverrigtingsmetings ontwikkel. Vantevore is die onsekerheid van ‘n enkele toetspunt gedurende ‘n pakkingsverrigtingstoets bereken. Die huidige studie brei geweldig uit hierop deur die onsekerheid van alle toetspunte te bereken en daarna hierdie onsekerheid te propageer, eerstens na die onsekerheid van die pakkingsverrigtingsvergelykings (bekend as regressie-onsekerheid) en daarna na die verrigtingsonsekerheid van die NTNKT. Daar word bepaal dat dit noodsaaklik is om gekorreleerde onsekerheid in die berekening van die regressie-onsekerheid in te sluit, maar dat dit weggelaat mag word, met relatiewe klein foute, by die NTNKT verrigtingsonsekerheid-vlak. Die ontwikkelde onsekerheidsanalisemetode voorsien streng regverdiging van enige verrigtings-verwante kommersiële besluite wat op pakkingsverrigtingstoetse gebaseer is. Die ontwikkelde pakkingsaangroeifaktor- en onsekerheidsanalisemetodes word toegepas op die weeklikse verrigtingstoetse van vier pakking-monsters wat oor nagenoeg 40 weke by ‘n pakkingsaangroeitoetsaanleg by ‘n steenkool-kragstasie in Suid-Afrika getoets is. Hierdie aanleg stel pakking-monsters bloot aan toestande naby aan dié in praktyk, insluitend die gebruik van kragstasie verkoelingswater. Alhoewel daar beperkings is op die toetsdata as gevolg van die relatiewe kort toetstydperk en lae neiging tot aangroei van die getoetsde pakkings, asook probleme wat die termiese toetsdata grootliks onbruikbaar maak, word die ontwikkelde metodes ten volle gedemonstreer.
Description
Thesis (MEng)--Stellenbosch University, 2020.
Keywords
Draught, Mechanical, Monte Carlo, Heat exchangers -- Fouling, Measurement uncertainty (Statistics), UCTD, Cooling towers and climate
Citation