Author's Copy PPCHEM ABSTRACT INTRODUCTION. PowerPlant Chemistry 2018, 20(3) Wolfgang Hater, Bill Smith, Paul McCann, and André de Bache

Transcription

1 Experience with the Application of a Film Forming Amine in the Connah's Quay Triple Stage Combined Cycle Gas Turbine Power Plant Operating in Cycling Mode Wolfgang Hater, Bill Smith, Paul McCann, and André de Bache ABSTRACT Due to the changing conditions of the energy market, many power plants have various periods of non-operation, ranging from a few days to months. Unprotected unit shutdown represents a serious corrosion risk and thus a risk for the integrity of key plant parts, such as the boiler or steam turbine. However, the established conservation methods of the water-steam cycle are not always applicable under the constraints of the modern power market, with unpredictable shutdown periods, while at the same time the plants have to remain available and may be required to run at short notice. Film forming amines (FFAs) offer excellent potential for the required flexible conservation process. The Uniper combined cycle gas turbine power plant located at Connah's Quay, UK, has assessed the applicability of FFAs for boiler and steam turbine protection. Besides a product based on a combination of FFAs with alkalising amines, a newly developed product containing solely the FFA was applied. Some key benefits could be demonstrated. The protection of the boiler and steam turbine could be achieved for a period of at least one month. The technology was able to protect all components of the watersteam cycle, including the areas of predominantly dry steam. Compared to dehumidification or nitrogen capping, minimal manpower was required for conservation. By the application of the newly developed product, the drawback of increased cationic conductivity levels was overcome, which remained close to the normal operation values. Due to the encouraging results, FFAs are now applied in all 4 units of the Connah's Quay power plant. INTRODUCTION Organic cycle chemistries based on film forming amines (FFAs) are increasingly being used as an alternative to conventional treatment programs for steam generators. Successful applications have been reported for both plants which are continuously operated [1 7] and plants under wet or dry lay-up [5,8,9]. The FFA molecule, often also referred to as polyamine or as fatty amine, adsorbs onto metal/metal oxide surfaces to form a hydrophobic film or barrier, which prevents corrosion by stopping water and other corrosive agents from contacting the metal/ metal oxide surface. Furthermore, the thin film fosters the formation of a smooth and compact iron oxide layer [10], which also plays an important role in preventing corrosion. Once formed, the protective film remains intact in both wet and dry conditions, even after dosing has stopped. This offers significant potential benefits for the preser vation of both drained and (partially) filled plants during shutdown, especially for plants under a cycling mode of operation. The technology of FFAs is now included in internationally accepted guidelines: The International Association for the Properties of Water and Steam (IAPWS) has published a Technical Guidance Document on the three main FFA molecules that have been the subject of intensive research and where significant application experience is available [11], which includes oleylpropylenediamine. Uniper UK's Connah's Quay combined cycle gas turbine (CCGT) power station has successfully trialled the use of film forming technology based on oleylpropylenediamine (OLDA) from Kurita (Cetamine ) for water-steam cycle preservation [12]. Over a three-year evaluation period, a comprehensive monitoring and site assessment programme was carried out. This paper reports on the experiences and results obtained. During inspections at Connah's Quay power plant, the internal surfaces of the heat recovery steam generator (HRSG) drums and low-pressure steam turbines were 2018 by Waesseri GmbH. All rights reserved. 136

2 investigated in situ for the presence of OLDA with a newly developed method [13]. Additionally, boiler tube samples were taken from the high-pressure evaporator and reheater stages from two different units of the power plant for destructive examination and laboratory analysis for the presence of OLDA. Connah's Quay Power Plant Connah's Quay Power Station (Figure 1) consists of four 355 MW single-shaft combined cycle units (Units 1 to 4). The HRSGs are vertical gas path drum-type boilers with three pressure stages (0.6, 3.6 and 12 MPa) and reheat. The final steam temperature is 540 C. The low-pressure superheater temperature is ca. 270 C. Make-up water is prepared by ion exchange and thermally and mechanically degassed in two steps in the deaerator. The station has a wet recirculating hybrid cooling system. Figure 2 shows a schematic of the water-steam cycle. The cycle chemistry is based on ammonia dosing to the feedwater to ph and sodium hydroxide dosing to the drums to achieve a ph between 9.2 and 9.4 in the high-pressure (HP) drum and between 9.5 and 9.8 in the intermediate-pressure (IP) and low-pressure (LP) drums. The station operating regime varies considerably, with between one and four units running on a daily start-up and shutdown basis. FFA technology was identified as a potential flexible preservation option for the plant due to difficulties establishing conventional preservation methods. Starting in December 2013, Cetamine V219 was dosed into the Unit 4 feedwater with reduced ammonia dosage and, starting in August 2015, Cetamine G850 was dosed into the Unit 1 feedwater in addition to ammonia. Both Cetamine products contain the same FFA molecule (OLDA). Cetamine G850 contains only OLDA, whereas Cetamine V219 additionally contains cyclohexylamine. After dosing was started, the targeted residual FFA concentration in the return condensate could be measured in Unit 4 after 420 hours of operation and in Unit 1 after 380 hours of operation. Residual FFA measurement was done photometrically with the bengalrose method [14]. Figure 1: The Uniper CCGT power plant in Connah's Quay, United Kingdom. 137

3 Exhaust CO PH DA EV Degasser HP EC1 IP/LP EC 1 Feedwater tank HP EC2 LP EV IP EC2 LP drum NaOH Ammonia Cetamine V219/-G850 IP EV IP drum NaOH SH HP IP LP G HP EC 3 SH Steam turbines HP drum NaOH Gas turbine HP EV SH RH G Condenser Figure 2: Flow scheme of the water-steam cycle of the Connah's Quay power plant. The areas of different pressure are marked by colour. DA deaerator EC economiser PH preheater EV evaporator SH superheater RH reheater G generator RESULTS General Water Chemistry The changeover from ammonia/naoh treatment did not show a significant influence on ph and direct conductivity. Steam Purity Conductivity after cation exchange (CACE) as well as degassed CACE (DCACE) in Unit 4 treated with OLDA and cyclohexylamine did not meet the quality requirement of steam purity to the steam turbines (Figure 3), whereas in Unit 1 (treated only with OLDA), the steam DCACE was generally below the required 0.2 µs cm 1 limit (Figure 4). However, CACE and DCACE were slightly increased in comparison to the previous treatment. In Figures 3 and 4, the cycling operation mode is reflected by intense sharp spikes in the conductivity readings at unit start-ups. The formation of acetate and formate was determined by ion chromatography. In Unit 4 dosed with OLDA and cyclohexylamine, the cause of the increased CACE was found to be carbon dioxide and also acetate and formate. The presence of acetate and formate was also reflected in 138

4 the DCACE measurement. This result is in accordance with other studies [15,4]. In contrast, the small elevation in CACE in the Unit 1 steam was almost all due to carbon dioxide. Figure 5 shows the average concentrations of acetate and formate for both units. The increased CACE in Unit 4 apparently results mainly from decomposition of the cyclohexylamine. This finding is confirmed by thermolysis studies under superheater conditions [16]. Figure 6 compares the individual contributions to the CACE in both units. In contrast to Unit 4, the contribution of low molec - ular acids is very low in Unit 1. There, the steam purity is maintained within the quality requirements for turbines. Plant Inspections During the trial period, plant inspections were carried out and various water-steam cycle surfaces tested for the presence of OLDA. Besides visual inspection and metallurgical studies, three methods of analysis were applied to determine OLDA on the system surfaces: the droplet test (test for hydrophobicity), the Kurita wipe test and also photoelectron spectroscopy (XPS). The droplet test is a simple and non-specific method to illustrate the wettability of a surface. The wipe test removes the FFA from the surface by wiping it with a solvent-soaked filter paper. It is Conductivity after Cation Exchange [µs cm 1 ] Start of Cetamine V219 dosing then transferred from the filter paper into an aqueous solution, and its presence in this solution is determined by the bengalrose photometric method. A visible pink colour or a significant absorbance from the bengalrose method is a clear proof of OLDA on the surface. The test itself and examples of its use in power plants are described in [17]. XPS provides the elemental composition of the upper surface layer as well as information on the bonding state of the elements. The information depth is ca. 10 nm. In contrast to the two other methods (droplet and wipe tests), XPS is generally destructive and cannot be carried out in situ. These methods are described in detail in [13]. In the following, the findings for selected parts of the water-steam cycle will be discussed. Drums All drums were very clean and completely free from organic deposits, such as gunk-balls. No active corrosion was observed. Loose iron deposits were significantly reduced; this was more pronounced in Unit 4, which can be attributed to the longer period of treatment with OLDA (Unit 4: 12 months; Unit 1: 3 months) at the time of this inspection. Figure 7 shows the appearance of the HP drum of Unit 4 and the LP drum of Unit 1. Degassed CACE CACE /01/13 01/09/14 04/19/14 07/28/14 11/05/14 02/13/15 05/24/15 09/01/15 12/10/15 Date [m/d/y] Figure 3: Conductivity after cation exchange (CACE) and degassed CACE (DCACE) in the reheat steam of Unit 4 treated with Cetamine V219 (OLDA and cyclohexylamine). The spikes mark unit start-ups. 139

5 2.0 Conductivity after Cation Exchange [µs cm 1 ] Start of Cetamine G850 dosing /15/15 07/25/15 08/04/15 08/14/15 08/24/15 09/03/15 09/13/15 Date [m/d/y] Figure 4: CACE in the superheated steam of Unit 1 treated with Cetamine G850 (OLDA). The spikes mark unit start-ups. Concentration [µg L 1 ] G850 Average V219 Average Acetate Formate Acetate Formate Acetate Formate HP Steam LP Steam IP Steam Conductivity after Cation Exchange [µs cm 1 ] CO 2 Acetate Formate Ionisation of pure water G850 V219 Unit 1 Unit 4 Figure 5: Average acetate and formate concentrations measured by ion chromatography in steam samples from Unit 4 (Cetamine V219) and Unit 1 (Cetamine G850). Figure 6: Average contributions to HP/IP steam CACE for Unit 4 (Cetamine V219) and Unit 1 (Cetamine G850). 140

6 Figure 7: Unit 4 HP drum after 12 months of treatment (left) and Unit 1 LP drum after 3 months of treatment (right) with film forming amines (OLDA). HP Evaporators Tube samples were taken from the Unit 4 HP evaporator for metallurgical analysis. The tube sample taken in 2015 after 2 years of treatment with OLDA showed only minor loose deposits. No active corrosion was seen. In comparison, the internal surfaces of a Unit 4 HP evaporator tube sample taken in 2012 before Cetamine dosing was started exhibited an outer layer of porous magnetite of ca. 100 µm thickness (Figure 8). This comparison showed that loose iron oxide had been gradually removed by the FFA. However, it did not remove the underlying dense protective magnetite layer. The surfaces of the tube samples were further studied by XPS. Nitrogen bonded to aliphatic organic carbon could be detected on all surfaces exposed to OLDA-based treatments, for both Unit 4 (Cetamine V219) and Unit 1 (Cetamine G850) in a concentration between 0.4 and 2.3 atom %, indicating the presence of OLDA on the tube surfaces. The tubes of Unit 4 showed a higher nitrogen concentration than Unit 1, again reflecting the longer treatment period with OLDA. Furthermore, a tube from Unit 4 from 2012 was analysed which had never been exposed to an OLDA-based treatment and no nitrogen was detected. Reheater Tube samples were taken from both units, and a typical internal appearance was found with no noticeable change following OLDA treatment (Figure 9). XPS detected nitrogen bonded to aliphatic carbon between 1.2 and 1.8 atom %, a strong indication of film formation also under dry steam conditions. Figure 8: Removal of loose iron oxide from Unit 4 HP evaporator waterside surfaces before (left) and after (right) 2 years of treatment with FFAs. 141

7 Figure 9: Reheater tubes from Unit 4 (left) and Unit 1 (right) after treatment with OLDA. No noticeable differences following the change of treatment. Figure 10: Appearance of Stage 5 (final stage) of the LP turbines of Unit 4 (left) and Unit 1 (right) after 12 and 3 months, respectively, of treatment with FFAs. LP Turbine Figure 10 shows pictures from the final stage (Stage 5) of the LP turbine of both units. Surfaces were very clean and showed clear hydrophobicity. No corrosion was observed. During a planned shutdown of Unit 1, all stages of the turbines could be assessed. All surfaces were clean and free from corrosion. Stages 1 3 had light magnetite deposits. Hydrophobicity was detected on Stages 4 and 5 only. During both inspections, a thorough evaluation of the accessible turbine stages using the Kurita wipe test was carried out. OLDA could clearly be measured on the LP turbine surfaces of both units. It was detected on all five stages of the Unit 1 LP turbine (only Stage 5 of Unit 4 could be accessed). Table 1 shows selected absorb ance readings corrected for the Blank. For comparison, the absorbance of the front side of a turbine blade that had not been exposed to an OLDA treatment is also shown. Hydrophobicity is apparently not an unambiguous indication of the presence of OLDA. 142

8 Sample Absorbance Untreated blade 0.05 Unit 4, Stage 5 Unit 1 Front side 0.25 Trailing edge 0.21 Disk and roots 0.28 Stage 5 Trailing surface 1.16 Inner front 0.33 Stage 3 Front face 0.29 Stage 1 Trailing face 0.23 Table 1: Presence of OLDA as determined by the Kurita wipe test on turbine surfaces treated with OLDA. For comparison, the value of an untreated blade is shown. CHANGING THE TREATMENT CONCEPT TO OLDA-BASED CHEMISTRY The results of the trials with OLDA-based treatment programmes showed an excellent preservation of the inspected plants for both units. Management and control of preservation with OLDA-based chemistry is simple. The treatment concept using Cetamine G850 containing only OLDA has the benefit of keeping the quality requirements of steam purity. start/stops of units, with unpredictable shutdown periods, while at the same time the plants have to remain available and may be required to run at short notice. Two products were applied: Cetamine V219, a combination of OLDA and cyclohexylamine, and Cetamine G850, containing only OLDA. The key findings can be summarised as follows: OLDA could be determined on both water and steam touched surfaces, including HP evaporators, reheaters and LP steam turbine cylinders. There was an evident internal cleaning effect in two units in the HP evaporator: loose iron oxides were removed, but not the magnetite layer. The reliability and maintenance effort of on-line chemical monitoring sensors were not affected by the film forming amine. Alkalising amines led to increased CACE caused by organic acids and carbon dioxide; treatment with the new Cetamine G850 provides steam in accordance with steam quality requirements. Components throughout the water-steam cycle were protected, including areas that could not be preserved by previous nitrogen capping or dehumidified air circulation. The preservation with film forming amine requires minimal manpower compared to preservation with nitrogen capping or dehumidification. There was no impact on unit availability or start-up times. The concept of FFA dosage for preservation and ammonia and sodium hydroxide for ph control has now been applied successfully for more than two years in all four units at Connah's Quay. Due to the overall positive findings, it was decided to adopt the supplementary treatment with Cetamine G850 in addition to conventional ammonia and NaOH condi - tioning for all four units at the Connah's Quay Power Station as of March The positive results have been confirmed by several plant inspections. The cleaning effect of OLDA observed in the HP evaporator of Unit 4 has also been seen in a second unit. Since 2017, Cetamine G850 has also been dosed into the auxiliary boilers (five Benson-type boilers with an overall 40 t per hour steam capacity), which are operated mainly during start-up and shutdown of the units. CONCLUSIONS The Uniper power station at Connah's Quay evaluated FFAs as a new treatment concept for meeting the preservation challenges of today's energy market, i.e. frequent REFERENCES [1] Allard, B., Chakraborti, S., Svensk Papperstidning 1983, 86(18), R 186. [2] Hater, W., Rudschützky, N., Olivet, D., PowerPlant Chemistry 2009, 11(2), 90. [3] Van Lier, R., Gerards, M., Savelkoul, J., VGB PowerTech 2012, 92(8), 84. [4] Kolander, B., de Bache, A., Hater, W., VGB PowerTech 2012, 92(8), 69. [5] Hater, W., Digiaro, C., Frayne, C., Proc., International Water Conference, 2012 (San Antonio, TX, U.S.A.). Engineers' Society of Western Pennsylvania, Pittsburgh, PA, U.S.A., Paper IWC [6] Hoock, B., Hater, W., de Bache, A., PowerPlant Chemistry 2015, 17(5),

9 [7] Sylwestrzak, E., Moszczynski, W., Hater, W., Dembowski, T., de Bache, A., VGB PowerTech 2016, 96(8), 69. [8] Hater, W., de Bache, A., Petrick, T., PowerPlant Chemistry 2014, 16(5), 284. [9] Wagner, R., Czempik, E., VGB PowerTech 2014, 94(3), 48. [10] Topp, H., Hater, W., de Bache, A., zum Kolk, C., PowerPlant Chemistry 2012, 14(1), 38. [11] Technical Guidance Document: Application of Film Forming Amines in Fossil, Combined Cycle, and Biomass Power Plants, International Associa - tion for the Properties of Water and Steam, IAPWS TGD8-16, available from [12] Smith, B., McCann, P., Hater, W., de Bache, A., Proc., VGB Conference "Chemistry in Power Plants", 2016 (Karlsruhe, Germany). VGB PowerTech, Essen, Germany, Paper #V07. [13] Smith, B., McCann, P., Mori, S., Uchida, K., Hater, W., Jasper, J., PowerPlant Chemistry 2017, 19(3), 129. [14] Stiller, K., Wittig, T., Urschey, M., PowerPlant Chemistry 2011, 13(10), 602. [15] Soellner, A., Glueck, W., Hoellger, K., Hater, W., de Bache, A., VGB PowerTech 2013, 93(3), 61. [16] Moed, D. H., Verliefde, A. R. D., Rietveld, L. C., Industrial and Engineering Chemistry Research 2015, 54(10), [17] Hater, W., Jasper, J., Disci-Zayed, D., Paper presented at the 2nd International Conference on Film Forming Substances, 2018 (Prague, Czech Republic). International Association for the Properties of Water and Steam. THE AUTHORS Wolfgang Hater (Ph.D., Physical Chemistry, Westphalian Wilhelms-University, Münster, Germany) started his professional career in 1989 at Henkel, where he worked for the industrial cleaner business. From 1993 until 2015 he worked in different technical positions in the water treatment divisions of Henkel and BK Giulini. Currently he is the technical director at Kurita Europe. He is the co-chairman of the working group Corrosion and Scale Inhibition of the European Federation of Corrosion and a member of the Power Cycle Chemistry working group of the International Association for the Properties of Water and Steam (IAPWS). Bill Smith, station chemist at Uniper's Connah's Quay Power Station, commenced working in electricity generation in the UK at Peterhead Power Station in 1984, followed by 10 years at a nuclear power plant at Heysham 2 Power Station. He has been a site chemist at Connah's Quay since Paul McCann (M.S., Chemistry, University of Nottingham, UK) is a specialist in power plant water-steam cycle chemistry, corrosion, and water treatment at the Uniper Technologies Ltd. global consulting unit in the UK. He has had 18 years of experience in the power industry since joining in Paul is vice-chair of the Power Cycle Chemistry working group of the International Association for the Properties of Water and Steam (IAPWS). He was also chair of the British and Irish Association for the Properties of Water and Steam (BIAPWS) from 2012 to André de Bache started his career at the Max Planck Institute for Bio-Inorganic Chemistry in Mülheim. In 1998 he joined Henkel, where he worked for the department of hygiene and microbiology. In 2004 he started working for BYK-Chemie in the application technology division of additives for plastics. In 2006 he joined Henkel Water Treatment, and started working in 2008 for ICL Water Solutions in product development for water conditioning, specializing in boiler water treatment. In 2012 he became the technical manager for boiler water additives, including traditional treatment products as well as products based on film forming amines. In 2014 he became the product manager for boiler water additives, filling this position since February 2015 for Kurita. CONTACT Wolfgang Hater Kurita Europe GmbH Niederheider Strasse Düsseldorf Germany Wolfgang.Hater@kurita.eu 144