Water piping networks may consist of a variety of pipe materials constructed over an extended period of time. Clay, asbestos-cement, ductile iron, steel, and more recently PVC, are common materials. Utilities are quite good at monitoring the condition of their network as a result of daily operations that offer a checkup on the network’s condition while crews repair leaks, install service connections, replace water meters, collect samples, or exercise fire hydrants. The situation gets more complex at residences and businesses.
Many homes and businesses have water piping constructed of copper pipes (many with lead solder joints) or galvanized steel pipe. Prior to the federal ban on lead pipe in 1986, service lines may have been constructed of lead, or galvanized pipe following a lead “pigtail”. These materials are all susceptible to corrosion from aggressive water, often in the form of pinholes or weakened joints. While a leaky pipe is a big problem, long-term consumption of leached contaminants is a more serious health risk.
The EPA Lead and Copper Rule, first promulgated in 1991, was revised in 2021, and EPA proposed additional improvements to the Rule in November 2023. Changes include an action level for lead of 10 ppb and 1.3 ppm for copper. In a major shift, concentrations are measured inside homes, businesses, and institutions (especially schools), not in the distribution system. This places utilities in the difficult position of being accountable for what happens inside homes, businesses, and institutions, even though they have no part in controlling the materials or usage of the water. Utilities must inventory their systems to identify service lines for replacement of lead and galvanized pipe (that meet certain requirements) within 10 years.
The Texas Commission on Environmental Quality (TCEQ) defines the degree of corrosiveness of a water based on a number of water quality parameters. Following the corrosive water crisis in Flint, Michigan, which started in 2014, TCEQ has taken a vigorous stance on utilities addressing corrosion. New sources are flagged as having a high potential for corrosion, and a corrosion study is required prior to approval.
A desktop corrosion study uses water quality data and other operational parameters to calculate multiple corrosive indices (Langelier Index, Ryznar Index, Aggressive Index). TCEQ establishes metrics for each index to determine if the source is corrosive, slightly corrosive, or non-corrosive. If deemed corrosive, the utility must implement a treatment scheme.
A pipe loop corrosion study is a more realistic assessment of corrosion potential. Samples of pipe from the distribution system, service lines, and residents are put together to construct a “pipe loop”. Water is circulated through the loop, and samples are taken after pre-determined residence times. Multiple loops allow for various waters to be tested: water directly from the system, and water treated after pH changes, or the addition of corrosion inhibitors. Tests are run for lead, copper, iron, and the various corrosion indices. Metal coupons can also be exposed to the tested water to measure the amount of physical material lost to corrosion.
Corrosive water must be addressed by using one of several treatment techniques.
- pH adjustment – Typically, pH is increased to alter the corrosivity of the water by adding sodium hydroxide or lime. Water can be aerated to remove carbon dioxide.
- Alkalinity adjustment – Alter corrosivity by increasing alkalinity with sodium bicarbonate or potash without increasing pH.
- Orthophosphate – Reduces lead solubility, copper and steel corrosion by building up a lead phosphate layer in pipes.
- Polyphosphate – Sequesters iron, manganese, and calcium in solution. May increase corrosion of lead and copper.
- Blended phosphates – Proprietary products that build up protective films.
Systems to feed these chemicals are fairly straight-forward, not capital intensive, and simple to operate. However, they must be continuously run, as protection will cease if dosing is stopped.
Testing prior to implementation is recommended, as results are very system-specific, and may take months for effects to be quantified.