In successful powerplant projects, either in the early project bidding or the engineering execution stage, before designing any water distribution, production or treatment system, you must know the water. The only way to fully understand the water type and its characteristics is with a thorough study of the water analysis report, says Wendy Wong, a senior process engineer at SNC-Lavalin.
You must recognize the reliable parameters, know which data need re-examination, and understand the water’s traits. Working early in the planning stages with good data from the water analysis report reduces engineering risks and ensures reliable water systems will support plant operations.
Water can be defined by its physical appearance (for example, color, suspended solids, and turbidity) and its biological and chemical properties. Water analyses should be performed by certified water laboratories. Be mindful that not all laboratories can perform all the tests required. This can be an issue when specific toxicity tests (such as for the surfactant Triton X-100) are needed, or when a minimum detection limit is required.
This was the case for a combined-cycle plant in Connecticut. A zinc detection level of two decimal points or better was required for the water supply to the cooling towers because the plant’s towers had a zinc limit for wastewater discharge of 1 mg/L or less.
Best practices to mitigate water collection challenges:
-
- Properly collect, sample, or preserve samples for accurate water quality to avoid plant design errors.
- Obtain sample bottles and preservatives from the laboratory conducting the required analyses.
- Use glass containers where oil or grease is present.
- Use amber-colored plastic containers (where use of plastic containers is permissible) to protect sample constituents, which can break down in sunlight.
- Pack samples in a cooler with ice for immediate shipping to the laboratory.
- Confirm maximum holding time with the lab to ensure representative results from the tests.
Another lesson learned: Sample residual chlorine, pH, and temperature at the sampling point with calibrated instruments to best reflect actual properties. Residual chlorine may be consumed by the living organisms in the water sample during shipping.
Water analysis report. To make sense of the water analysis report, it’s important to know the water source that feeds the sampling points, and where and when the samples were taken. All the physical characteristics should be examined together to provide an overview on corrosivity, particle’s physical size, scaling potential, and fouling tendency of the water. The biological and chemical properties indicate the choice of chlorination, biological organism activities, dissolved salts content, scaling potential, fouling potential, reactivity, salinity, and toxicity of the water.
Missing parameters. Many powerplants in the US and Canada use a city water supply. Unfortunately, only a few parameters are required to be reported to the public—such as residual chlorine, conductivity, coliform, etc. So even if the water supply comes from a city’s potable water network, which is a good and clean water source, use of a water-treatment-system design based on a limited water analysis can spell disaster.
For example, the system may not satisfy power requirements under all operating conditions, perform the various operations needed, or hold to certain parameter limits in the wastewater discharge permit because the concentrations of some constituents in the water supply are unknown.
Quality variance. Seasonal and weather-related changes and facility production schedules (for grey and wastewater) can impact water quality. While drastic changes in water temperature occur between the cold and warm months, not checking water temperature onsite can create problems, especially when designing the biological treatment, cooling water, and reverse osmosis (RO) systems, or any treatment process that depends on water temperature.
To illustrate: The capacity of high-pressure RO pumps/motors may be insufficient when operating in winter if the design temperature is based on warmer samples taken in summer. The design team should obtain at least one water sample per season or at varied days/times to ensure representative samples are collected.
The design engineer also should consider weather-pattern changes and the future impacts to water quality when designing a new plant and using surface-water data 10 years or older. They may not be representative of the current water condition. New land developments in the area can impact surface water analysis as well.
Worst-case scenario. The design’s technical specifications in a request for proposal sometimes summarize historical water analyses and provide only the worst-case scenario based on the peak value of each constituent without providing the individual water analysis. Result: Water-treatment bidders likely will offer larger or more complex equipment than necessary, and it will cost more.
Also, when the source water’s characteristics and trends are not fully understood, unnecessary equipment and more complex, high-maintenance systems could be added to a new plant and they may produce more wastewater.
When design engineers have sufficient water data and recognize the trends, they can design a plant to accommodate the worst-case scenario with minimum impacts to costs. Whether by recirculation, blending, providing temporary storage, or operating the standby water-treatment trains, they can provide sustainable, cost-effective solutions.
Other general guidelines. A water analysis report can be unreliable. Typos, use of the wrong units of measurement or confusing nomenclature (for example, mg/L versus mg/L as CaCO3), expired samples, or simply the wrong analysis can result in incorrect reports.
Guidelines circulated in the industry for verifying the reliability or plausibility of a water analysis report that you may find helpful are provided below. Listed in decreasing order of reliability, they were published in 1991 by G Solt under Dewplan (WT) Ltd, which later became part of Veolia Water.
1. Natural water from the streams, ocean and underground where pH is around neutral, the water shall be chemically balanced—that is, the sum of the cations equals the sum of anions.
2. Conductivity should roughly equal the total cations multiplied by 1.6 if in ppm CaCO3. If a total-dissolved-solids (TDS) value is given, it should equal the total cations, assuming equivalent weight of between 50 to 70.
3. Bicarbonate, HCO3, is often about the same as the calcium ions, Ca2+. If it is more than the total hardness (Ca2+ and Mg2+), be suspicious.
4. Ca2+ concentration is generally between four to eight times the Mg2+.
5. In the absence of reliable data, Na+ and Cl– concentrations can be assumed the same, especially if either of them is high.
In sum, designing water distribution, production and treatment systems requires reliable and accurate data. Bid requests with limited parameters or insufficient water analyses may bring in proposals with big cost differences from water treatment vendors based on assumptions and their experience with the water source, with some proposals possibly being double the cost of others.
Taking the time to work early in the planning stages with experienced engineers, studying the water trends, and defining sampling parameters and frequency pays off. The good, reliable, and representative water analyses become the foundation of an adaptable and sustainable system to support long-term powerplant operation for both normal and upset water supply.
Wendy Wong, PE, a senior process engineer at SNC-Lavalin, has 26 years of global experience in chemical and water-treatment processes for the power, oil and gas, and pharmaceutical industries. Her expertise includes treatment of cooling water and demineralized water, plus seawater desalination and cycle chemistry for boilers.
