Nowadays, performance studies on the amperometric total residual oxidant (TRO) sensor are only in the bench test stage and have not been conducted under specific maritime conditions with Ballast Water Management System (BWMS). In this study, the application of the amperometric TRO sensor in land-based biological efficacy (BE) testing, operation and maintenance (O&M) testing, as well as shipboard (SB) testing, was explored by comparing with the existing di-phenylene-diamine (DPD) TRO sensor. The results showed that the average TRO measurement deviation between the amperometric sensor and the DPD sensor was within ±10% in valid BE test cycles and the O&M testing exceeding 47 operating hours. The TRO value measured by amperometric sensor exhibited significant fluctuations, but the improved control logic could achieve smoothing out the fluctuation, with stability comparable to or even higher than that of the DPD sensor. The practicality and reliability of the amperometric sensor in electrochlorination-based BWMS were further verified through SB testing.

  • TRO measurement deviation with the amperometric sensor was within an acceptable range.

  • TRO signal output with the amperometric sensor was affected by salinity.

  • Smoothing control logic could mitigate the measurement fluctuation caused by the TRO sensor.

  • The reliability of the amperometric sensor was further verified through shipboard testing.

Ballast water is used to maintain the ship's stability and maneuverability and control the trim and heel of the hull (Thach & Phan 2023). However, ballast water contains various exclusive organisms from both ballast and de-ballast regions, recognized as the major vector for the invasion of alien organisms potentially (Duan et al. 2023). It is estimated that 7,000–10,000 different species globally every day, such as marine microorganisms, plants and animals, are transferred through the ballast water pathway (Lakshmi et al. 2021). The arbitrary discharge of untreated ballast water can inevitably cause ecosystem disorder and lead to significant economic losses (Marbuah et al. 2014; Kurniawan et al. 2022).

In order to reduce the risk of alien species introductions, the International Maritime Organization (IMO) adopted the International Convention for the Control and Management of Ships' Ballast Water and Sediments (referred to as the Convention) in 2004. The Convention entered into force on September 8, 2017, with clear limits on the discharge concentration of living organisms in ballast water (Regulation D-2) (IMO 2004). In addition to the IMO, the United States also promulgated a Ballast Water Discharge Standard (BWDS) for its territorial waters in 2012 and the final rule of the United States Coast Guard (USCG) entered into force on the same year (USCG 2012). Therefore, Ballast Water Management System (BWMS) become necessary for international ships' navigation (Sayinli et al. 2022).

BWMS, based on electrochlorination, are widely adopted in the shipping field to meet regulation D-2 of its higher inactivation efficiency (Cha et al. 2015). According to statistics, approximately 40% of BWMS evaluated by IMO use electrochlorination technology and electrochlorination BWMS account for approximately one-third of BWMS-type approved by USCG (Zimmer-Faust et al. 2014; USCG 2024). Given that ballast water treatment can generate ecotoxic total residual oxidants (TROs) potentially, both the Group of Experts on Scientific Aspects of Marine Pollution (GESAMP-BWWG) of IMO and the United States Environmental Protection Agency (USEPA) propose requirements for monitoring TRO in BWMS (USEPA 2013; GESAMP-BWWG 2015–2022; Kim et al. 2016). Therefore, accurate TRO measurement is crucial for BWMS process control.

Currently, TRO measurement mainly includes the di-phenylene-diamine (DPD) method and the amperometric method. TRO sensors based on the DPD method employ colorimetric assays utilizing chemical reagents. Total oxidants in the sample water react with these reagents and the higher the oxidant concentration, the darker the color (Buchan et al. 2005; ISO 2017). Although TRO sensors based on the DPD method were already type-approved due to their high accuracy and stability and widely used in electrochlorination-based BWMS, there are also many issues in actual operation (Zimmer-Faust et al. 2014; Zhou et al. 2020). For example, the DPD TRO sensor is usually installed online and an elaborate sampling line is arranged for TRO measurement. Additionally, reagents are indispensable for the measurement and should be replaced regularly within months. The DPD reagent is active after preparation and should be kept at a low temperature, which makes it more challenging on board ships with higher ambient temperatures. Sample drain after analysis needs to be led back to the main ballast line or neutralized before discharge, due to residual oxidant in the sampling water.

Amperometric TRO sensors are essentially electrochemical, producing a current proportional to chlorine concentration (Zimmer-Faust et al. 2014). Compared with DPD TRO sensors, amperometric TRO sensors are reagent-free and can be installed in line with direct insertion into the main ballast line, which greatly reduces economic costs and installation difficulties. Nowadays, the amperometric TRO sensor with three electrodes, including a platinum working electrode (WE), a platinum auxiliary electrode/counter electrode (CE) and a silver/silver chloride reference electrode (RE), has been successfully developed. This solved the problems of frequent calibration, low measurement accuracy and unstable output signal caused by double-electrode TRO sensors (Li et al. 2022). However, the performance characteristics of amperometric TRO sensors are only at the research stage and cannot be directly applied to BWMS. Differences in factors, such as seawater quality, ship environment and durability, can significantly affect the TRO sensor's performance (Badalyan et al. 2009; Zimmer-Faust et al. 2014). Therefore, it is vital to conduct a comprehensive performance evaluation of the amperometric TRO sensor with BWMS.

This study aimed to verify the measurement accuracy and robustness of the amperometric TRO sensor in comparison with the DPD TRO sensor. Studies included land-based (LB) testing in three salinity ranges (freshwater, brackish water and marine water), operation and maintenance testing and shipboard (SB) testing. The results could guide the application of the amperometric TRO sensor in electrochlorination-based BWMS.

LB biological efficacy testing

The biological efficacy (BE) test was conducted independently by a test facility in Hundested, Denmark. The BE test applied challenge water in compliance with IMO BWMS Code and Environmental Technology Verification (ETV) protocol (USEPA, 2010; IMO 2018). The challenge water was delivered from the source tank to the treated tank and approximately 200 m3 of challenge water was treated by side-stream electrolytic BWMS with a flow rate of 300 m3/h and a TRO dosage target of 7.5 mg/L. An online DPD sensor and an in-line amperometric sensor were installed adjacently on the main ballast line to measure the TRO level. The installation diagram is shown in Figure 1. DC for electrolysis was regulated in accordance with the TRO level measured by the DPD sensor. Three test cycles in each salinity (freshwater, brackish and marine water) were carried out independently by the test facility. Physical–chemical characteristics of challenge water for BE testing are shown in Table 1.
Table 1

Physical–chemical characteristics of challenge water for land-based testing

Test cycleSalinity (PSU)Temperature (°C)pHTSS (mg/L)POC (mg/L)DOC (mg/L)Duration (min)
F-1 0.38 14 8.3 68 5.7 7.3 40 
F-2 0.38 20 7.8 66 7.1 6.5 40 
F-3 0.40 20 7.7 60 6.8 6.0 40 
B-1 (not calibrated) 15 9.9 8.1 60 8.0 6.6 40 
B-2 18 15 8.4 67 7.8 6.9 40 
B-3 18 17 8.2 71 7.6 6.9 40 
M-1 (not recorded) 29 17 8.1 42 5.9 7.2 40 
M-2 28 19 8.4 52 7.6 7.2 40 
M-3 28 19 8.3 47 7.0 7.4 40 
O&M-1 (not calibrated) – – – – – – 900 
O&M-2 18 13 8.2 – – – 960 
O&M-3 15 18 7.8 – – – 120 
O&M-4 18 17 7.5 – – – 480 
O&M-5 17 18 7.6 – – – 360 
Test cycleSalinity (PSU)Temperature (°C)pHTSS (mg/L)POC (mg/L)DOC (mg/L)Duration (min)
F-1 0.38 14 8.3 68 5.7 7.3 40 
F-2 0.38 20 7.8 66 7.1 6.5 40 
F-3 0.40 20 7.7 60 6.8 6.0 40 
B-1 (not calibrated) 15 9.9 8.1 60 8.0 6.6 40 
B-2 18 15 8.4 67 7.8 6.9 40 
B-3 18 17 8.2 71 7.6 6.9 40 
M-1 (not recorded) 29 17 8.1 42 5.9 7.2 40 
M-2 28 19 8.4 52 7.6 7.2 40 
M-3 28 19 8.3 47 7.0 7.4 40 
O&M-1 (not calibrated) – – – – – – 900 
O&M-2 18 13 8.2 – – – 960 
O&M-3 15 18 7.8 – – – 120 
O&M-4 18 17 7.5 – – – 480 
O&M-5 17 18 7.6 – – – 360 
Figure 1

TRO sensors’ installation diagram.

Figure 1

TRO sensors’ installation diagram.

Close modal

Operation and maintenance testing

To verify the reliability of the TRO sensors incorporated with BWMS, the operation and maintenance (O&M) test was conducted independently by a test facility that had the same system as BE testing and arranged amid the BE test cycles. The O&M test was operated the same as the BE test, but the ballast pump takes brackish water locally from the Hundested harbor in Hundested, Denmark. The treated water will be discharged directly into the harbor after the neutralization at the end of the discharging piping. Physical–chemical characteristics of challenge water for O&M testing are shown in Table 1. The sensors were operated for a total of 53 h exceeding the 50 h as specified by the ETV protocol during five O&M tests (USEPA 2010).

Shipboard testing

The SB testing was conducted on board a 300,000 DWT very large crude carrier (VLCC) which was installed with two identical sets of BWMS with treatment rated capacity (TRC) of 3,200 m3/h per system. An online DPD sensor and an in-line amperometric sensor were installed adjacently on the main ballast line of the port side to measure the TRO level during the ballasting treatment. The electrolysis unit takes up water from the sea chest in the engine room after being filtered by a 50-μm automatic back-flushing filter. Chlorine was produced by electrolysis and injected into the main ballast line in the pump room by the dosing pump. Two TRO sampling ports were arranged downstream of the injection point to measure the TRO concentration after fully mixing. DC for electrolysis was regulated in accordance with the TRO level measured by the DPD sensor to reach the dosing target of 7.5 mg/L TRO.

TRO control logic updating for the amperometric sensor during ballasting

The control logic responses for TRO changes during the ballasting process were updated as follows.

  • The output reading of the TRO level was the average value of the previous output TRO level and the present actual TRO level.

  • The output reading would remain the same as the previous output reading if the present actual TRO level deviated from the previous output TRO level beyond the pre-set deviation ratio (>20%). The output reading would be the average value of the previous output TRO level and present the actual TRO level if the two consecutive actual TRO levels were beyond the pre-set deviation ratio.

The control loop for the regulation of current input to the electrolyser remained the same with the DPD TRO sensor.

Analytical methods

Temperature, salinity and PH were measured by in-line probes and operational flow was measured by an in-line flow meter. Chemical parameters of total suspended solids (TSS), dissolved organic carbon (DOC) and particulate organic carbon (POC) were sampled and analyzed independently by the test facility for compliance with challenge water requirements prescribed in BWMS Code and ETV. A programmable logic controller (PLC, Mitsubishi, Q series) was used for the control of the whole system and data logging. All the TRO measurements from the DPD sensor and the amperometric sensor were automatically recorded every minute during the whole testing. The data were extracted from the PLC and saved as CSV files for statistical analysis.

Measurement deviation comparison during LB testing

As a critical component of BWMS, TRO sensors must be tested and verified by an independent laboratory before type approval. The amperometric TRO sensor used in the testing was operated simultaneously with the DPD sensor during nine LB tests and five O&M tests. Excluding the unreliable test results due to the lack of calibration of the amperometric sensor in B-1 and O&M-1 tests, an invalid test of M-1 test, a total of 11 tests could be calculated for the comparison of TRO measurement deviation. All the test results are shown in Figure 2. TRO measurement deviations were 7.9% (B-2), 17.5% (B-3), −3.4% (M-2), 8.0% (M-3), −0.6% (F-1), −4.1% (F-2), −2.7% (F-3), 5.2% (O&M-2), 5.4% (O&M-3), 4.8% (O&M-4) and 6.8% (O&M-5), respectively. The results indicated that the acceptable average measurement deviation between the DPD sensor and the amperometric sensor was within ±10% in 10 out of 11 valid test cycles, except the B-3 test with a deviation of 17.5%. The TRO signal output from the amperometric sensor in marine and brackish water was usually higher than that of the DPD sensor (positive deviation), although the deviation was less than 10%. Throughout the entire O&M test cycle from O&M-2 to O&M-5, the average TRO value measured by the amperometric sensor was comparable to that of the DPD sensor, with a deviation of only 4.8–6.6%. It could be concluded that the system demonstrated satisfactory operation in natural seawater during more than 47 h. This result also confirmed that the three-electrode amperometric sensor combined with the RE could overcome the impact of electrode passivation caused by continuous redox reactions (Campo et al. 2005; Li et al. 2022). In freshwater testing from F-1 to F-2, the average TRO value measured by the amperometric sensor was similar to or slightly lower than that of the DPD sensor, with a deviation from −0.6 to −4.1%. TRO typically exists as a mixture of chlorine, hypochlorous acid, hypochlorite, etc (Li et al. 2023). The ratio of different chlorides can affect the signal output from the amperometric sensor (Seymour et al. 2020). Compared with the positive deviation observed in seawater, the negative deviation in freshwater might be due to the different TRO consumption rates caused by the water characteristics at the test facility, resulting in the chloride in the test water being mainly hypochlorite, thus reducing the signal output from the amperometric sensor. Based on the above results, the measurement deviation between the DPD sensor and the amperometric sensor was within ±10% in valid LB test cycles, except the B-3 test cycle. Salinity affected the measurement of the amperometric sensor, and measurement accuracy declined with higher salinity. The reliable operation during all valid O&M tests also validated the suitability of the amperometric sensor throughout the whole ballasting process on board ships.
Figure 2

Average TRO levels measured with the DPD sensor and amperometric sensor in LB testing.

Figure 2

Average TRO levels measured with the DPD sensor and amperometric sensor in LB testing.

Close modal

Measurement fluctuation evaluation during LB testing

More stable TRO measurements facilitate more stable adjustment of the electrolysis current output by the rectifier, controlling produced chlorine concentration, thereby ensuring that TRO after mixing in the main ballast pipeline remains within the target value range (Zimmer-Faust et al. 2014; Zhang & Wang 2022). To evaluate the stability of TRO measurements, test data of TRO measurement for each water quality were extracted and the fluctuation was analyzed, including F-2, B-2, M-2 and O&M-2 tests, as shown in Figure 3. The TRO control target of BWMS was based on a DPD sensor, with TRO levels fluctuating around 7.5 mg/L. The TRO variation trend measured by the amperometric sensor was similar to that of the DPD sensor; however, due to the interference of electroactive species in water on the chlorine reduction current signal (Zhou et al. 2020), more drastic fluctuation was inevitably observed in all the tests for the amperometric sensor, which might cause significant deviation from the target TRO level in a short period (e.g., B-3 test cycle).
Figure 3

TRO levels measured under continuous operation in different test cycles.

Figure 3

TRO levels measured under continuous operation in different test cycles.

Close modal

Additionally, amperometric sensors were susceptible to interference from pH levels. As is well known, due to the sensitivity of amperometric sensors to measuring the content of hypochlorous acid, a pH of 5.0 to 7.0 is the ideal operation range for amperometric sensors. When the pH was greater than 7.0, hypochlorite ion became the main free chlorine, with hypochlorous acid significantly reduced, which would inevitably affect the accuracy of the amperometric probe (Malkov et al. 2009). Studies reported that organic chloramines could also affect the measurement of amperometric sensors (Jensen & Johnson 1990). To reduce the dependency of amperometric sensors on pH, most manufacturers typically employ internal pH compensation to improve the potential accuracy of amperometric sensors. However, due to the complexity of seawater quality, internal pH compensation also had its limitations. Temperature changes were also a major factor affecting the measurement of amperometric sensors (Zimmer-Faust et al. 2014; Li 2021). Studies indicated that three main areas affected by temperature were the membrane permeability rate, the conductivity of the electrolyte and the sample pH (Malkov et al. 2009). Although the amperometric sensor was equipped with a temperature compensation function, due to potential differences in membrane materials, electrode geometry, thermistor position, etc., no temperature compensation algorithm could accurately reflect all changes in the water matrix or the response of chlorine to these changes, inevitably causing measurement fluctuations.

In order to smooth the measurement fluctuation of the amperometric TRO sensor, this study updated the control logic of the existing sensor according to the description in Chapter 2.4 and demonstrated mitigated TRO measurement fluctuation by comparing the actual and processed TRO reading of the amperometric sensor in O&M-2 test cycle, as shown in Figure 4. As can be seen, the average value of processed reading remained the same as the original data, at 7.89 mg/L, indicating that this control logic improvement did not significantly affect the accuracy of the instrument. However, the standard deviation (SD) decreased from 0.94 mg/L for the actual reading to 0.56 mg/L for the processed reading, which was more comparable to the SD of 0.48 mg/L with the DPD TRO sensor. This decline in SD showed less fluctuation of measurement and more stable reading after processing, which further benefited the output current adjustment from the rectifier.
Figure 4

Actual and processed TRO reading by amperometric sensor in O&M-2 test cycle.

Figure 4

Actual and processed TRO reading by amperometric sensor in O&M-2 test cycle.

Close modal

SB testing verification

Figure 5 shows the TRO levels measured during SB testing before and after updating the control logic to verify the practicality and reliability of the amperometric sensor in complex environments. In the test before updating control logic, the average TRO value measured by the amperometric sensor was 8.18 mg/L, which resulted in a 6.1% deviation compared to that by the DPD sensor (7.71 mg/L). The average TRO values measured by the amperometric sensor and DPD sensor in the updated test were 7.57 mg/L and 7.06 mg/L, respectively, with a deviation of 6.7%. It could be seen that the measurement deviations between the DPD sensor and the amperometric sensor in two SB tests were both within an acceptable range of 7% and measurement deviation remained similar before and after the control logic update, which indicated that the control logic change did not affect the accuracy of the sensor.
Figure 5

TRO levels in SB testing before (a) and after (b) control logic update.

Figure 5

TRO levels in SB testing before (a) and after (b) control logic update.

Close modal

The variation trend of the amperometric sensor was similar to that of the DPD sensor, but more drastic fluctuation was observed in a test without updated control logic, which could cause significant instantaneous deviations. To further evaluate this fluctuation, this study conducted an SD analysis on the SB test results before updating the logic. The results showed that the SD value (2.15 mg/L) generated by the amperometric sensor during the measurement was much higher than that by the DPD sensor (1.62 mg/L). This fluctuation was also observed in LB and O&M testing. With the update of control logic, the fluctuation of the amperometric sensor had been significantly mitigated, as shown in Figure 5(b). The SD value decreased sharply from 2.15 mg/L to 1.33 mg/L, even lower than the SD (1.58 mg/L) of the DPD sensor, which indicated that the updated logic assisted in maintaining a more stable TRO measurement with the amperometric sensor.

The following conclusions could be drawn from this study:

  • The acceptable average measurement deviation between the DPD sensor and amperometric sensor was within ±10% in 10 out of 11 valid LB test cycles. Salinity affected the measurement of the amperometric sensor, and measurement accuracy declined with higher salinity.

  • The system demonstrated satisfactory operation in natural seawater with no augmentations during the O&M testing period of more than 47 h, which confirmed that this amperometric sensor could overcome the impact of electrode passivation and met the applicability conditions.

  • More drastic fluctuation was inevitably observed in all the tests for the amperometric TRO sensor, although the measurement deviation was below 10%.

  • Smoothing the control logic of the amperometric TRO sensor could significantly reduce measurement fluctuations (about 40%), avoiding rapid changes in current input (power supply) to the electrolyser.

This study was supported by the National Quality Infrastructure (2022YFF0610403).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.

Badalyan
A.
,
Buff
J.
,
Holmes
M.
,
Chow
C. W. K.
&
Vitanage
D.
(
2009
)
On-line free-chlorine/total-chlorine monitors’ evaluation – a step towards a correct choice of residual disinfectant monitor
,
Aqua
,
58
(
3
), 181–190.
Buchan
K. A. H.
,
Martin-Robichaud
D. J.
&
Benfey
T. J.
(
2005
)
Measurement of dissolved ozone in sea water: a comparison of methods
,
Aquacultural Engineering
,
33
(
3
),
225
231
.
Campo
F. J. D.
,
Ordeig
O.
&
Muñoz
F. J.
(
2005
)
Improved free chlorine amperometric sensor chip for drinking water applications
,
Analytica Chimica Acta
,
554
(
1–2
),
98
104
.
Cha
H. G.
,
Seo
M. H.
,
Lee
H. Y.
,
Lee
J. H.
,
Lee
D. S.
,
Shin
K.
&
Choi
K. H.
(
2015
)
Enhancing the efficacy of electrolytic chlorination for ballast water treatment by adding carbon dioxide
,
Marine Pollution Bulletin
,
95
(
1
),
315
323
.
Duan
D.
,
Xu
F.
,
Wang
T.
&
Fu
H.
(
2023
)
The effect of filtration and electrolysis on ballast water treatment
,
Ocean Engineering
,
268
, 113301.
GESAMP-BWWG
(
2015–2022
)
Report of the GESAMP-Ballast Water Working Group
.
London, UK
:
International Maritime Organization
.
IMO
(
2004
)
International Convention for the Control and Management of Ship's Ballast Water and Sediments
.
London, UK
:
International Maritime Organization
.
IMO
(
2018
)
Code for Approval of Ballast Water Management Systems (BWMS Code)
,
Vol. Resolution MEPC.300(72)
.
London, UK
:
International Maritime Organization
.
ISO
(
2017
)
Water Quality – Determination of Free Chlorine and Total Chlorine – Part 2: Colorimetric Method Using N,N-Dialkyl-1,4-Phenylenediamine, for Routine Control Purposes
.
Geneva, Switzerland
:
International Organization for Standardization
.
Kim
E. C.
,
Oh
J. H.
&
Lee
S. G.
(
2016
)
Consideration on the maximum allowable dosage of active substances produced by ballast water management system using electrolysis
,
International Journal of e-Navigation and Maritime Economy
,
4
(
C
),
88
96
.
Kurniawan
S. B.
,
Pambudi
D. S. A.
,
Ahmad
M. M.
,
Alfanda
B. D.
,
Imron
M. F.
&
Abdullah
S. R. S.
(
2022
)
Ecological impacts of ballast water loading and discharge: insight into the toxicity and accumulation of disinfection by-products
,
Heliyon
,
8
(
3
),
e09107
.
Lakshmi
E.
,
Priya
M.
&
Achari
V. S.
(
2021
)
An overview on the treatment of ballast water in ships
,
Ocean & Coastal Management
,
199
,
105296
.
Li
T.
,
Wang
Z.
,
Wang
C.
,
Huang
J.
&
Zhou
M.
(
2022
)
Chlorination in the pandemic times: the current state of the art for monitoring chlorine residual in water and chlorine exposure in air
,
Science of the Total Environment
,
838
(
Pt 3
),
156193
.
Malkov
V. B.
,
Zachman
B.
&
Scribner
T.
(
2009
)
Comparison of on-line chlorine analysis methods and instrumentation built on amperometric and colorimetric technologies
,
Water Quality Technology Conference and Exposition
,
2009
,
30
51
.
Marbuah
G.
,
Gren
I. M.
&
Mckie
B.
(
2014
)
Economics of harmful invasive species: a review
,
Diversity
,
6
,
500
523
.
Sayinli
B.
,
Dong
Y.
,
Park
Y.
,
Bhatnagar
A.
&
Sillanpaa
M.
(
2022
)
Recent progress and challenges facing ballast water treatment – a review
,
Chemosphere
,
291
(
Pt 2
),
132776
.
Seymour
I.
,
O'Sullivan
B.
,
Lovera
P.
,
Rohan
J. F.
&
O'Riordan
A.
(
2020
)
Electrochemical detection of free-chlorine in water samples facilitated by in-situ pH control using interdigitated microelectrodes
,
Sensors and Actuators B: Chemical
,
325
, 128774.
Thach
N. D.
&
Phan
V. H.
(
2023
)
Development of UV reactor controller in ballast water treatment system to minimize the biological threat on marine environment
,
Journal of Sea Research
, 198,
102465
.
USCG
(
2012
)
Standards for Living Organisms in Ships’ Ballast Water Discharged in U.S. Waters
.
Washington, DC, US
:
United States Coast Guard
.
USCG
(
2024
)
BWMS Type Approval Certificates
.
Washington, DC, US
:
United States Coast Guard
.
USEPA
(
2010
)
Generic Protocol for the Verification of Ballast Water Treatment Technology (ETV Protocol)
,
Vol. EPA/600/R-10/146
.
Washington, DC, US
:
United States Environmental Protection Agency
.
USEPA
(
2013
)
Vessel General Permit
.
Washington, DC, US
:
United States Environmental Protection Agency
.
Zhang
R.-x.
&
Wang
Y.-l.
(
2022
)
A method for monitoring residual chlorine in ship ballast water by energy conversion
,
Ship & Ocean Engineering
,
51
(
04
),
90
94
.
Zhou
M.
,
Li
T.
,
Zu
M.
,
Zhang
S.
&
Zhao
H.
(
2020
)
Membrane-based colorimetric flow-injection system for online free chlorine monitoring in drinking water
,
Sensors and Actuators B Chemical
,
327
, 128905.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).